A. (fclfcj INVESTIGATION GLYCOCOLL AND SOME OF ITS PRODUCTS OF DECOMPOSITION EBEN N. HORSFORD, A.M., Rumford Proffegsor in the University at Cambridge. INVESTIGATION GLYCOCOLL AND SOME OF ITS PRODUCTS OF DECOMPOSITION. EBEN N. HORSFORD, A.M., Rumford Professor in the University at Cambridge. 3/102T EXTRACTED FROM THE AMERICAN JOURNAL OF SCIENCE AND ARTS, VOLS. III. & IT, SECOND SERIES, 1847. NEW HAVEN: PRINTED BY B. L. HAMLEN, Printer to Yale College. 1847. I To JOHN W. WEBSTER, M.D., ERVING PROFESSOR IN THE UNIVERSITY AT CAMBRIDGE: THE FOLLOWING PAPER, IN TESTIMONY OF GRATEFUL REMEMBRANCE, IS RESPECTFULLY DEDICATED, BY THE AUTHOR. • INVESTIGATION, &c. As there are laws in physics, whose evolution, carefully traced, would constitute a general history of this department of science, so there are bodies in chemistry, whose career, if we may employ such an expression, accurately followed out, would acquaint us with the prominent periods through which this science has passed. The history of sulphuric acid, for example, may almost be said to be the history of technical chemistry, as that of hydrosulphuric acid and ammonia is of analytical; or, as that of oxygen is of theoretical chemistry. Among the bodies belonging to organic chemistry, perhaps no one presents itself, whose history, fairly exhibited from the time of its discovery down to this date, contains more obvious and lasting impressions of the periods through which this branch of chemical science has passed, than the sugar of gelatine. It bears the stamp of a period when an organic analysis had scarcely been made; of a time when the substance was not obtained pure; then of a time when, from the perfection of methods and appa- ratus, analyses enjoyed the whole confidence of the chemist, and it was only necessary that formulas should coincide with them, in order to their being considered correct; and finally, of a time in which the relations to other substances came in to share in the solution of the problem of the constitution of an organic body. These circumstances may justify a glance at the history of this interesting body. Braconnot,* in 1820, by treating isinglass with sulphuric acid, obtained a body of sweet taste, ready solubility in water, difficult solubility in alcohol, capable of uniting with nitric acid, and in this state of combination uniting with alkalies and alkaline earths; to which substance he gave the name of sugar of gela- tine, (sucre de gelatine.) Boussingault,f to whom we are indebted for the first analysis of this body, gave it the formula C^H^^O,,, which, ex- pressed in equivalents, isC]2HlsiN3On. From this body he obtained a crystallizable compound, with the protoxyd of lead, which, upon analysis, yielded the formula C, 2 H, 3i N3 07 +3PbO. Three atoms of protoxyd of lead had taken the place of two atoms of water. * Annal. de Chim. et de Phys., T. xiii, p. 113. t L'Institut, No. 245. Phar. Cent. Blatt, No. 50,1838. 2 6 Prof. E. N. Horsford on Glycocoll, If this be regarded neutral, Boussingault remarks, the body without water would be C4 H4jlN03. The investigation undertaken at this period (1838) was re- sumed in 1840, '41 * when, from the same chemist, a new for- mula, obviously based upon a conscientious trust in the results of analysis, was produced. Meanwhile, Mulderf had obtained the same body with Leucin, by treating glue with caustic potash. His analysis led him to the formula C8 H9 N2 07. With protoxyd of lead it lost two atoms of water. Its composition would then have been C8 H7 N2 05 4-2HO, the two atoms of water being replaceable by two atoms of protoxyd of lead. This formula differs from that of Boussingault, chiefly in that it is about two-thirds as large. In that of Boussingault, an equal loss of water was, however, re- placed by three atoms of protoxyd of lead, instead of two. With the gelatine-sugar-nitric acid, [acide-nitro-saccharique; leim-zucker-salpetersaure,) Mulder obtained a compound with baryta, of unusual constitution, which he expressed by the fol- lowing formula— (BaO, C8 H7 N2 05)+2(BaO,N05). His formula for gelatine-sugar-nitric acid, is C,H11N40I,=CBHIIN,0,+2N05+2HOl while that of Boussingault, for this compound, was C4H6iN2O10=C4H4iNO3+NO5-}-2HO. These differ from each other in the relative and absolute quan- tities of hydrogen and oxygen, and yet not so widely but that the want of correspondence might be attributed to slight im- purity of substance. Boussingault analyzed a compound of this acid with oxyd of copper, of the following constitution :— C4H6iNoO10+2CuO = (C4H4iN03,NOs)+(2CuO,HO). At a temperature of 165° C, [329° F.] this compound lost 17T1 per cent, of water, leaving C4H2iNO,NOs+2CuO. Deducting the two atoms of oxide of copper, there remains a body— C4H2iNO. In this condition of the question, as to the constitution of gly- cocoll, the investigation was resumed by Boussingault. He anal- yzed the body itself, and several most interesting compounds of it with oxyds of copper, lead and silver, nitric acid and nitrates of metallic oxyds. From these he derived the formula CI6H15N40)t+3HO, * Ann. de Chim. et de Phys., 3d Ser., T. i, p. 257-270. t Nat. en Sclieik. arcliief, 1838, p. 146. and some of its Products of Decomposition. 7 in which the three atoms of water were replaced with four atoms of metallic oxyd. The gelatine-sugar-nitric acid he found to be CieH15N40II>4NOJ+9HO, and its compounds with bases to be C10H15N4OI,+4NO.,+4(8)MO4-2aq._ This exceedingly complicated formula, and its high atomic weight, together with the fact that the several formulas for gela- tine sugar, enumerated above, though differing absolutely from each other, are nevertheless but slight modifications from once, twice, three and four times the following formula, C4H4N03, led naturally to the conclusion that the differences might be attrib- uted partly to the impurity of the substances analyzed, and partly to the imperfect atomic weights at that time in use. Beyond the description and analysis of the body itself, and the few salts above enumerated, gelatine sugar had met with no de- tailed examinations. These considerations gave occasion to the investigation which follows. It is scarcely necessary to add to the statement of its having been conducted at Giessen, that the counsel and cooperation there enjoyed, have united with the recol- lections of this labor, some of the most grateful memories of a life. Formula of Glycocoll. Finding it impossible to obtain as large a per-centage of carbon by the combustion of gelatine sugar, either in chromate of lead or oxyd of copper, as had been found by both Mulder and Bous- singault, no conclusion remained, but that the analyses they had recorded, and those we had made, were of different bodies. After the analyses of several compounds of this body with hydrochloric, sulphuric and nitric acids, oxyd of copper, nitrate of silver, and bisulphate of potash, the conviction was established that its con- stitution was C 4 H 4 NO 3, to which in crystallized gelatine sugar, an atom of water is united.* Upon comparing the per-centages derived from this formula with the results of analysis in Boussingault's last investigation, the differences will be seen to be scarcely greater than frequently occur in a series of the best determinations. The body analyzed by Boussingault, was dried at 120° O, [248° F.] That analyzed by us lost nothing in weight at 150° C, * For the anhydrous body, we have adopted the already proposed name Gly- cocoll. The inappropriatenessof the name sugar, has been noticed by Dessaigne. The attribute of sweetness it shares with AqO, S2 °2» °xide of glyceryl and ni- trous ether, and beside, it is incapable of fermentaiion. We submit the symbol Gl., as least likely to lead to confusion. G. might con- found with Glycium, (Glucinum, Beryllium,) and Gly. is already appropriated to Glycerile. 8 Prof. E. N. Horsford on Glycocoll, [304° F.] When exposed upon a watch crystal to the heat of a lamp, with a metallic screen between, and at such distance that the escape of vapor is barely discernible, a part of the mass in contact with the glass becomes browned, while another portion melts and shoots into crystals. These continue to form even after other portions have become charred. Rubbed together with finely pulverized hydrate of baryta, it becomes almost instantly fluid, the whole dissolving readily in water, from which, in pro- cess of time, crystals containing baryta and glycocoll deposit themselves. Here, in the act of combination, the water from one or both the ingredients was given up. The above circumstances, and others yet to be noted, induced the opinion that at a certain temperature, a lower one longer con- tinued, an atom of water from one half of the hydrated glycocoll might be given up, and the remaining half take its place. This would give almost precisely the analytical results of both Mulder and Boussingault, and yield from the formula determined on by us, precisely the composition given by Mulder. 2(C4 H4 N03, HO)—HO=C8 H9 N? 07. Below, follow the estimated per cent, constitution, according to the above formula, and the average of a series of analyses by Boussingault and Mulder. Estimate. Bous. Mulder 8 equiv. Carbon, .... 9 " Hydrogen, ... 2 " Nitrogen, 7 " Oxygen, .... 3404 638 19-85 39-73 33 79 644 19-90 39-70 3417 6-49 1984 39-50 10000 10000 10000 The effort to expel this half atom of water was unsuccessful. A temperature of 150° C. [302° F.] produced scarcely a percep- tive diminution in weight. At 170° C. [338° F.] it began to brown with the escape of gaseous products of decomposition. At 190° C. [374° F.] though portions had become quite charred, others had merely melted and crystallized anew. The support which the analyses of Boussingault give to the formula C.H.NO^HO, will justify the following juxtaposition of the estimated per cents. and the actual results. Glycocoll and Oxyd of Silver ; dried at 110° C. [230° F.] Boussingault. horsford. Ci6Hi5N4011+4AgO. C4H4N03 + AgO. Carbon, . Hydrogen, Nitrogen, Oxygen, Ox. silver, Equiv. 16 = 96 15 = 15 4 = 56 11 = 88 4 == 464 Estimate. Result. Equiv. FstiniaiP. 1318 219 7-69 13-21 63 73 10000 13-35 2-08 7-78 12-96 64-53 13-62 221 8-07 12-35 6375 4 = 24 4 = 4 1 = 14 3 = 24 1 = 116 719 100-00 100-00 | 182 and some of its Products of Decomposition. 9 Glycocoll and Oxyd of Copper; dried at 120° C. [243° F.] Boussingault. Horsford. C16H15N4011+CuO. C4H4N03+CuO. Carbon, . Hydrogen, Nitrogen, Oxygen, Ox. copper, Equiv. 16 = 96 15 = 15 4 = 56 11 = 88 4 = 158-8 Estimate. Result. Equiv. Estimate 23-19 3-62 13-53 21-29 38-37 101-00 2357 3-75 37 60 4 = 24 4=4 1 = 14 3 = 24 1 = 39-7 105-7 22-68 3-77 13-24 2270 37-61 10000 413-8 Glycocoll and Oxyd of Lead ; dried at 120°. [248° F.] Boussingault. Horsford. C1GH15N4011+4PbO. C4H4N03+PbO. Carbon, . Hydrogen, Nitrogen, Oxygen, Ox. lead, Equiv. , Estimate. Result. Equiv. Estimate. 13-48 224 7-86 13-50 62-92 16 = 96 15 = 15 4 = 56 11 = 83 4 = 446-8 701-8 13-67 213 7-97 12-57 63-66 13-58 212 7-78 11-62 64-90 4 = 24 4=4 1 = 14 3 = 24 1 = 111-7 10000 100-00 177-7 100-00 Glycocoll and Nitric Acid, dried in vacuo by ordinary temperature. Boussingault. Horsiord. C16H Carbon, Hydrogen, . Nitrogen, . Oxygen, . 15N4011-f4N05+9HO. C4H4NOs, HO+N05, HO. Equiv. Estimate. 17-39 4-34 20-29 57-98 100 00 Result. Equiv. — 4~=—24 6=6 2 = 28 10 = 80 138 Estimate. 16 = 96 24 = 24 8 = 112 40 = 320 17-32 4-53 20-23 57-92 100-00 17-39 4-34 20-29 57-98 552 10000 At 100° C. [212° F.] this salt lost 9-18 per cent., and at 110° C. [230° F.] 4-5 per cent, more, with which it began to brown. This loss, 13-68 per cent., corresponds nearly with 13-77 per cent., the water in the salt. Its formula would then be C4H4N03,N05. Glycocoll and Nitrate of Silver. Boussingault. Horsford. C16H15N4011+2HO+4(AgO,N05). C4H4N03 + AgO, NOfi. Carbon, Hydrogen, . Nitrogen, . Oxygen, . Ox. silver, . Equiv. 16 = 96 17 = 17 8 = 112 33 = 264 4 = 464 Estimate. Result 10-09 1-84 11-50 27-87 48-70 Equiv. Estimate. 10-07 1-78 11-75 27-72 48-68 4 = 24 4=4 2 = 28 8 = 64 1 =116 1016 1-69 11-87 26 75 49-53 953 100-00 100-00 236 100-00 Glycocoll and Nitrate of Potash. Boussingau It. o5). Horsford. C16H15T C4H4N03KO. NOv Carbon, Equiv. 16 =~96~ Estimate. —14 14~ Result. Equiv. | Estimate. 14-26 4 = 24 14-35 Hydrogen, . 17 = 17 255 2-42 4=4 2-39 Nitrogen, . 8 = 112 16-50 2 = 28 16-78 Oxygen, . 33 = 264 38-86 8 = 64 38-21 Potash, 4 = 189-6 27-95 27-49 1 = 49-4 167-4 28-27 678-6 100-00 100-00 10 Prof. E. N. Horsford on Glycocoll, Glycocoll and Nitrate of Copper. _________ Boussingault. Ci6H15N4011+4N05+8CuO-r-9HO. Equiv. Estimate. Result. Carbon, 16 = 96 11 02 11-04 Hydrogen, 24 = 24 2-75 2-89 Nitrogen, . 8 = 112 12-86 1208 Oxygen, . 40 = 320 36-79 37-45 Ox. copper, 8 = 317-6 ttt>9-0 36-58 10000 36 54 100-00 If we review the estimates from the two systems of formulas, and the results of analysis recorded above, it will be seen— 1st. That the analyses of the compounds with nitrate of silver, more nearly correspond with the estimates from Boussingault's formula. 2d. That the analyses of the compounds with nitric acid, ni- trate of potash and nitrate of copper, are suited equally well to either formula. And 3d. That the analyses of the remaining compounds, viz : with the oxyds of silver, copper and lead, correspond more nearly with the simpler formula. As additional reasons for adopting the simpler formula, the fol- lowing, drawn from our investigation, may be recorded. 4th. The analyses give this constitution: as the average of four combustions for carbon and hydrogen, and two for nitrogen, will show, placed side by side with the estimates from the formula C4H4N03,HO. Theory. Experiment. Hydrogen, .... Nitrogen, .... Oxygen, .... 4 equiv. = 24 5 " = 5 1 " = 14 4 " =32 32-00 6-67 18-67 4266 100-00 3198 6-87 18-79 42-36 75 100 00 5th. It forms a compound in which sulphuric acid replaces the atom of water: C4 H4 N03, SOs corresponding with Boussingault's nitrate, dried at 110° C. [230° F.]: C4H4N03,N05. 6th. It forms a compound with oxyd of copper, of this formula: C4H4N03, CuO, HO, which, at 100° C, [212° F.] loses an atom of water. 7th. It forms a compound in which the atom of water is re- placed by an atom of nitrate of silver : C4H4N03; AgO,N05. 8th. It forms, when long digested with sulphuric acid, a salt of this constitution: C4H4N0,,S03H0> H4Nd,S03H0. 5 Horsford. c4h4no3,ho|C;;o^05 Equiv. Estimate. 4 = 24 6=6 2 = 28 10 = 80 2 = 79-4 11 02 2-75 12 86 36-79 36-58 217-4 100-00 and some of its Products of Decomposition. 11 The ammonia is taken from the hydrated glycocoll, which may be considered as fumarate of ammonia: C4H4N03,HO=H4NO,C4H03. 9th. It decomposes when subjected to the action ol the gal- vanic battery; an acid (fumaric?) and oxygen appearing at one pole, and an alkali (ammonia) and hydrogen appearing at the other. 10th. It may be derived from hippuric acid by treatment with a mineral acid, in which case a neutral salt of glycocoll is formed and benzoic acid set free : or by treatment with potash, in which case benzoate of potash is formed and glycocoll is set free. Hippuric acid, - - C13H9N06 Benzoic acid, - - - C14Hs03 Glycocoll, - - - C4 H4N03 We may add also that the formulae of Boussingault and Mul- der have not the advantage of so great simplicity. Preparation of Glycocoll. The recent brilliant discovery of Dessaigne* that by boiling hippuric acid with sulphuric, hydrochloric or any of the stronger acids, this body separates into benzoic acid and a salt of glycocoll, rendered the preparation of the latter, in purity, a task of no dif- ficulty. It was, of course, necessary first of all to obtain a quan- tity of hippuric acid. This acid was prepared according to the method proposed by Dr. Bensch,f by evaporating in a water bath, the morning urine of stall-fed horses, to from one-eighth to one-tenth of its vol- ume ; adding hydrochloric acid till all effervescence ceased; set- ting aside, in a cool place, for the perfect separation of the hip- puric acid; filtering through linen, and pressing; dissolving in freshly prepared solution of hydrate of lime, with addition ot boiling water; filtering, as before, and pressing ; heating the fil- trate to boiling, acidifying with solution of alum, cooling to 40° C. [104° F.], adding solution of bicarbonate of soda till no farther precipitation takes place; filtering and pressing ; and precipitation of the filtrate with hydrochloric acid. After washing, filtering and pressing the precipitate of hippuric acid, it was again dissolved in boiling water, and blood coal added in the proportion of half an ounce to a pound of moist acid, again filtered at boiling heat through paper, and set aside to crystallize. By this method pris- matic crystals are obtained of the most perfect whiteness, an inch in length, and from one to two lines in diameter. * Compt. Rend., xxi, p. 1224-1227. Liebig's Annalen, Bd., lviii, S. 322. \ Liebig's Annalen, Bd., lviii, S. 267. 12 Prof. E. N. Horsford on Glycocoll, From three to four ounces of hippuric acid were digested in a flask of one litre* capacity, over a spirit lamp, in four times their weight of concentrated hydrochloric acid, until entirely dissolved. A larger quantity is less manageable, and the subsequent treat- ment less expeditious. It is well to continue a gentle heat, with the addition of water half an hour after the solution is completed. In this time a part of the benzoic acid gathers into oily drops and sinks to the bottom, becoming, as the whole cools, a solid crys- talline mass, not easily removed from the flask. The presence of water retains the hydrochlorate of glycocoll in solution. After the whole mass has become cool, the benzoic acid having for the most part crystallized out, the contents of the flask are poured upon the filter and washed till the filtrate no longer tastes sour. The filtrate is then evaporated in an open porcelain basin over a water bath, nearly to dryness,—by which process the excess of hydrochloric acid and benzoic acid are, for the most part, removed. Redissolving and again evaporating two or three times repeated, at length give the hydrochlorate of glycocoll quite pure. Am- monia is then added to the syrup till it yields an alkaline reaction. Absolute alcohol being then added, the remaining traces of benzoic acid and the salammoniac are dissolved, while the glycocoll is thrown down in the form of a white precipitate, consisting of myriads of minute prismatic crystals. As the body is slightly soluble in alcohol, and more so in alcohol containing salammo- niac, it is well to let the fluid stand a few hours. The precipi- tate is then brought upon a filter and washed with absolute alco- hol, till the filtrate no longer gives a precipitate with nitrate of silver. Properties of Glycocoll. Thus obtained, hydrate of glycocoll tastes sweet, though less so than cane sugar,—has neither acid nor alkaline reaction; dis- solves in from 4-24 to 4-35 parts of water; is more soluble in hot than in cold spirits of wine ; is quite insoluble in ether, and scarcely less in absolute alcohol. When heated with a concentrated solution of caustic potash, in excess, it assumes, with the evolution of ammonia, a fine bril- liant red color. If the heat be continued, the color gradually dis- appears. Heated with hydrate of baryta or oxyd of lead, the same bril- liant color is produced. With sulphate of copper, a trace of glycocoll prevents the pre- cipitation by potash, and the solution assumes a characteristic blue color. Boiled with oxyd of copper or its hydrate, it yields the same blue solution, which, if concentrated, crystallizes in fine needles. # = 0-2201 English gallon. and some of its Products of Decomposition. 13 With nitrate of suboxyd of mercury, it gives a precipitate of metallic mercury. From a concentrated solution in diluted spirits of wine or in water, in process of time large prismatic crystals are formed, which apparently belong to the monoclinate system, of the combi- nation 30 P. OP. +P. x Px . Prof. Kopp, to whom I am indebted for an examination of these crystals and of several others of salts of glycocoll, obtained from crystals prepared by Prof. v. Liebig, an admeasurement of the sharper angles of x P, (through which the orthodiagonal passes,) giving 6G£°. The combustions for carbon and hydrogen were made with chromate of lead ; those for nitrogen, according to the method of Varrentrapp and Will,— I. 0-6770 grm. gave 0-7940 carbonic acid and 0-4170 water. II. III. IV. V. VI. 0-5.57(5 0-4670 0-4003 0-1338 0-1937 VI. " 0-6607 " " " 0-3474 " 0-5455 " " " 0-2882 « 0-4686 " " " 0-2478 " 0-4100 platin-salammoniac. " 0-5646 " " These determinations correspond in per cent, with I. II. III. iv. v. Carbon, 31-89 32-31 31-81 31-92 . . Hydrogen, 6S4 6-92 6 85 6-87 Nitrogen, . . . . . . From these may be derived the formula, C4H4N03,HO, as will be seen by comparing the estimated and average actual per cents, of the several elements. 19-24 18-36 Theory. Experiment. Carbon, ... Hydrogen, ... Nitrogen, ... Oxygen, 4 equiv. = 24 5 " = 5 1 " =14 4 " =32 3200 6-66 18-66 42 68 31-98 6-87 18-79 42 36 75 ioo-oo 10000 The atomic weight of glycocoll is, with the above constitu- tion, 66. Glycocoll and Hydrochloric Acid. Neutral Hydrochlorate of Glycocoll. Gl, HOI, HO. This is the product of boiling hippuric acid with concentrated hydrochloric acid, as already described. If the filtrate, page 376, be carefully evaporated to syrup consistence, and suffered quietly to cool, the whole mass becomes filled with groups of long, flat prisms, perfectly transparent, and of the greatest brilliancy. The mother liquor poured off, and the crystals washed with spirits of wine, gives the salt in the utmost purity. A second and third 3 14 Prof. E. N. Horsford on Glycocoll, portion of crystals may be obtained by concentrating the mother liquor and similar treatment. ; This salt slowly absorbs moisture from the air and deliquesces : over sulphuric acid, the crystals retain their form and constitution any length of time. It tastes sour, and slightly but positively stiptic ; reacts acid : dissolves readily in water; in hot spirits of wine, and slightly in absolute alcohol. The substance dried over sulphuric acid, on combustion with chromate of lead, I. 0-2368 grm. gave 0-1833 carbonic acid and 0-1272 water. II. III. IV. V. VI. 0-3218 0-2853 1-5920 1-0008 1-5300 " 0-2555 " 01711 " 0 5698 platin-salammoniac. 2-0562 chlorid of silver. 1-2961 " " 1-9300 " " in. IV. VI. Expressed in per cents., the above determinations correspond with i. Carbon, 21-11 Hydrogen, 5-96 Nitrogen, . . Chlorine. . . n. 21-28 5-95 . . 12-57...... .....31-91 31-99 31-99 With these numbers, the following formula is in accordance: C4H4N03,H01, HO, as the annexed comparison will show. Theory. Experiment. Carbon, ... 4 equiv = 24 21-42 21-2') Hydrogen, ... 6 " = 6 5-35 5-95 Nitrogen, ... 1 " = 14 12-56 12 57 Oxvgen, - 4 " = 32 28-53 2840 Chlorine, ... 1 " = 35-4 3214 31-94 111-4 10000 ? 100 00 Basic Hydrochlorate of Glycocoll. (a.) 2G1, HC1, HO. This salt is formed by adding hydrochloric acid to a cold con- centrated solution of glycocoll in water, and then pouring in al- cohol till the solution becomes slightly turbid. Crystals of the utmost transiency and regularity commence forming imme- diately. To continue the crystallization, alcohol is added drop by drop, at intervals. It is perhaps better to let the solution in water slowly evaporate over sulphuric acid. In this way larger crystals, of surpassing beauty, were obtained. They are rhombic prisms of 87° and 93°. They do not deliquesce like those of the neutral salt, upon exposure to the air. The salt has a pleasant sour, and at the same time sweet taste, reminding one of fine fresh pippins. The solution reddens litmus with chromate of lead. and some of its Products of Decomposition. 15 I. 0-3505 grm. gave 0-3550 carbonic acid and 0-1743 water. II. 0-2758 " " 0-6729 platin-salammoniac. III. 1-5050 " " 1-1940 chlorid of silver. These determinations expressed in per cents.: i. n. in. Carbon, 27-59 Hydrogen, 5-52 . . . . Nitrogen, . . 15-37 Chlorine, . . . . Correspond with the formula C4 H4 N03, HC1+C4 H4 NO: which, calculated, gives: 19-58 HO, Theory. 27-05 5-63 15-7H 31-59 19-95 Experiment. 27 59 5-52 15-37 31-94 19-58 TOO-00 Carbon, ... Hydrogen, -Nitrogen, ... Oxygen, ... Chlorine, 8 equiv 10 " 2 " 7 " 1 " = 48 = 10 = 28 = 56 = 35-4 1774 Basic Hydrochlorate of Glycocoll. (b.). 2(G1, HO) + HCl. This salt was obtained by dissolving glycocoll in hydrochloric acid and leaving the solution to a quiet crystallization. The ex- act proportions of acid and base necessary to procuring it have not been determined. Indeed, it will appear obvious, after the ac- counts of the hydrochloric and sulphuric acid compounds, that the task of accurately fixing the temperature, concentration and quantity of the several ingredients necessary to the formation of a given compound of acid, glycocoll and water, will be exceed- ingly difficult. Combustion with chromate of lead gave the following results: I. 1-0290 grm. gave 0-9840 carbonic acid and 0-5580 water. II. 1-0890 " " 0-8180 chlorid of silver. III. 0.9760 " " 0-7305 " " IV. 0-9710 " "■ 0-7290 " In per cents, expressed, i. n. Carbon, 26-08 . . Hydrogen, 6-02 . . Chlorine, . . 18-53 These numbers conduct to the formula, 2(c;h4 no3, ho)+hci, the estimated per cents, of which with the results of analysis are here placed side by side. 18-46 IV. 18-41 ... Theory. Experiment. Carbon, -Hydrogen, Nitrogen, Oxygen, Chlorine, 8 equiv. = 48 11 « =11 2 " = 28 8 " =64 1 « = 35-4 26-4(>' 6 06 15-43 33-12 1891 100-00 26-08 6-02 18-47 10000 186-4 16 Prof. E. N. Horsford on Glycocoll, The rational constitution of this salt may be considered as one atom of hydrated glycocoll, united to one atom of hydrochlorate of hydrated glycocoll, thus, Gl, HO + G1, HC1, HO. Basic Hydrochlorate of Glycocoll. (c.) 3G1, 2HC1, 2HO. This salt was prepared in the same manner as the last; a sim- ple solution of glycocoll in hydrochloric acid, set aside to crystal- lize. The acid was, however, in excess. It was also prepared by passing dry hydrochloric acid gas over melted hydrate of glycocoll. For this purpose, a gramme and a half of substance was distributed along the bottom of a Liebig's drying apparatus, and carefully heated with a spirit lamp; at the same time conducting over it hydrochloric acid gas.' At a tem- perature of between 150° C. and 170° C. [between 302° F. and 338° F.] as determined in an oil bath, the glycocoll melts in the acid atmosphere. It was found better, however, to employ the simple lamp. With the latter the apparatus could be readily inclined or half inverted, to spread the molten substance over the interior of the tube, and thus facilitate absorption. The absorption is attended with the escape of aqueous vapor. The process was continued until no further increase in weight was observed. At each interval the hydrochloric acid was thor- oughly removed, by long continued passing of dry air through the tube before weighing. At the end of the absorption, the glycocoll usually became slightly green, owing doubtless to a trace of de- composition and separation of carbon. From an analysis of crystals obtained by the first method, I. 1-2560 grm. of substance, gave 1-2520 grm. chlorid of silver. By the other method, II. 1-9727 grm. of hydrated glycocoll, increased in weight to 2-4580 grm. Precipitated with silver, this gave 2-3855 grm. chlo- rid of silver. The increase in weight was 24-60 per cent. The per centages of chlorine, I. 24-59 ; II. 24-23. These numbers correspond to the formula, 3(C4 H4 N03) + HC1 + 2H0, which requires 24-51 per cent, of chlorine. Basic Hydrochlorate of Glycocoll. (d.) 3G1, 2HC1, HO. This salt is prepared precisely as the last mentioned, both by crystallization from the acid solution and by leading dry hydro- chloric acid gas over fused hydrate of glycocoll. The notice of this salt would scarcely have been ventured up- on, had not a precisely corresponding compound with sulphuric acid been analyzed. It will contribute to show how multifarious may be the relations of a body, that combines as a salt, and yet possesses both acid and basic properties. and some of its Products of Decomposition. 17 From crystals of the salt prepared as above mentioned, I. 1-2864 grm. gave 1-3203 grm. chlorid of silver. II. By leading dry hydrochloric acid gas over glycocoll in the manner already described, fc compound was formed, of which, 1-1370 grm. gave 1-1845 grm. chlorid silver. In per cent, expressed, these determinations give, i. n. Chlorine, . . . 25-43 25-72 corresponding with the formula 3(C4 H4N03) + 2HC1+H0, which requires 25-30 per cent, of chlorine. As the probable ra- tional constitution of the above salt the following is submitted, (Gl HC1 + G1, HO) + Gl HC1. Anhydrous Hydrochlorate of Glycocoll. Having found a basic hydrochlorate, which might be regarded as a double salt of one atom of hydrate of glycocoll, with one atom of anhydrous hydrochlorate of glycocoll: C4 H4 N03, HO + C4 H4 N03, HC1, and especially having found as will be seen below, an anhydrous sulphate of glycocoll, it was natural to suppose that the anhydrous hydrochlorate might be obtained by itself, viz. C4 H4 N03, HC1. To this end absolute alcohol was saturated with hydrochloric acid gas, and this added to a solution of glycocoll in hot spirits of wine. Upon evaporating the liquid, delicate prismatic crystals appeared which deliquesced with the greatest rapidity. They even dissolved in absolute alcohol. This latter circumstance led to the supposition that the crystals might have been a double salt of hydrate of glycocoll with hydrochlorate of oxide of ethyle. This supposition was further strengthened from an analysis of a sulphate of corresponding constitution soon to be noticed. Compounds of glycocoll and sulphuric acid are even more re- markable than those with hydrochloric acid. As little success at- tended the effort to ascertain the precise conditions under which some of them are formed, as rewarded the labors with the com- pounds already described. Of these, two, the double sulphate of glycocoll and oxyd of ammonium, and the anhydrous sulphate of glycocoll, have especial interest, as they throw much light over the constitution and nature of this body. Anhydrous Sulphate of Glycocoll. Gl, S03. By dissolving glycocoll in hot spirits of wine, cooling, adding sulphuric acid drop by drop, and setting aside in a quiet place, after a day or two there are formed beautiful elongated thin flat prisms with right angled terminal planes. From another portion the salt crystallized in the most delicate attenuated tables of the 18 Prof E. N. Horsford on Glycocoll, greatest brilliancy. It is soluble in water and hot diluted alcohol, and quite insoluble in absolute alcohol and ether. It tastes sour and reddens litmus paper, does not change upon exposure to the air, and loses no weight by 100° C. (212° F.) Combustion with chromate of lead gave the following results: I. 0-5147 grm. gave 0-4257 carbonic acid and 0-2509 water. II. 0-3134 " " 0-2574 " " " 01616 " 0-1260 " " 0-7039 platin-salammoniac. sul- III. IV. V. 0-3134 01541 0-3397 0-4248 phate of baryta. In per cent, expressed agreeing with, with chlorid of barium 0-4673 grm. i. 22-55 5-41 Carbon, Hydrogen Nitrogen, . . Sulphuric acid, . . Which give the formula C n. 22-40 5-72 in. 22-30 IV. 1305 37-97 H, NO,, SO. as the comparison of estimated and analytical results shows. Theory. Experiment. Hydrogen, .... Nitrogen, .... Oxygen, .... Sulphuric acid, ... 4 equiv. = 24 4 " = 4 1 " =14 3 " =24 1 " =40 22-66 377 13 20 22 64 37-73 22 42 5-56 1305 21-00 37-97 ] 0000 106 10000 Repeated combustions did not enable us to lessen the percent- age of hydrogen. The variation from the theory is, doubtless, to be attributed to the absorption in the chlorid of calcium tube, of a small quantity of sulphurous acid, which escaped from the combustion tube. This explanation unfortunately occurred after repeated analyses had consumed the stock of salt. This constitution is remarkable in the field of organic chem- istry. On its borders we have a similar instance in anhydrous sulphate of ammonia, NH3-4-S03. Sulphate of Hydrate of Glycocoll. GL S03, HO. This salt was obtained from a solution similarly prepared to that which yielded the anhydrous salt, except that the solution was boiled with sulphuric acid, instead of the latter being added to the cold solution. It crystallizes in short prisms, reminding one of sulphate of copper, and the crystals, though small, are of ex- ceeding beauty and perfection of form. They do not change upon exposure to the ah. A single determination only was made, and that of the nitro- gen. The other determinations were not made, from want of and some of its Products of Decomposition. 19 substance, all subsequent efforts to form the salt having failed. By Varrentrapp and Will's method:— 0-3367 grm. gave 0-2943 grm. platin-salammoniac. In per cent, expressed, Nitrogen 12-37. This corresponds with the formula C4 H4 N03, S03, HO, which requires 12-17 per cent, of nitrogen. Basic Sulphate of Glycocoll. (a.) 3G1 HO, 2S03, 2HO. If to a solution of glycocoll in diluted spirits of wine, sulphu- ric acid in excess be added, and set aside, in twenty-four hours long rectangular prismatic crystals form upon the bottom of the containing vessel. A very considerable excess of sulphuric acid did not change the constitution of the crystals. They taste and react acid, and like the salts already described suffer nothing from exposure to the air. Combustion with chromate of lead gave the following results: I. 04199 grm. gave 03528 carbonic acid and 0-2149 water. II. 0-3944 " " 0-3219 " " " 01974 " HI. 0-2399 " " by Varrentrapp and Will's method, 0-5067 grm. platin-salammoniac. IV. 0-6866 grm. gave 0-4928 sulphate of baryta. V. 0-5808 " " 0-4170 " " VI. 0-4532 " " 0-3225 " " VII. 0-4960 " " 0-3500 " " In per cent, expressed the above determinations correspond with i. n. in. iv. v. vi. vn. Carbon, 22-91 22-25.......... Hydrogen, 5-68 5-56.......... Nitrogen, . . . . 13-31........ Sulphuric acid,......24-62 24-62 24 20.24-40 These numbers give the formula, 3(C4 H4 NO3, HO)+2(S03, HO). The juxtaposition of the estimated per cents, and analytical results follows:— Theory. Experiment. 22-58 562 13-31 3403 24 46 Hydrogen, .... Nitrogen, .... Oxygen, .... Sulphuric acid, - 12 equiv. = 72 17 " = 17 3 " =42 14 " =112 2 " =80 22-29 5-26 1300 34-69 24-76 323 10000 10000 The following formula is submitted as expressing the rational constitution of this salt:— (Gl HO, S03 HO+G1 HO)+(Gl HO, S03 HO). The following sulphuric acid compounds were none of them completely analyzed. They were prepared in small portions 20 Prof E. N. Horsford on Glycocoll, while seeking to obtain a neutral sulphate of hydrate of glyco- coll; and it was-not until the capacity of this body to combine with others of such different nature, and in such varied propor- tions became fully apparent, that the existence of so complex and unusual compounds was believed. The crystallized salts were for the most part groups of elon- gated prisms. Basic Sulphate of Glycocoll. (b.) 3G1, 2S03,HO. The constitution of this salt differs from that of the preceding in the amount of water. As both of them were dried in the air over sulphuric acid, and suffered no change, this difference is attrib- utable doubtless to the degree of concentration, or difference of temperature. It will be observed that it corresponds precisely with a basic hydrochlorate (d), whose constitution is given on page 16. Willi chlorid of barium, 0-2182 grm. of crystals, gave 0-1940 grm. sulphate of baryta. In per cent, sulphuric acid 27-74. This corresponds with the formula 3(C4 H4 N03)+2S03+HO, which requires 2787 parts in 100. As the probable rational con- stitution of this salt the following is submitted : (Gl, S03+Gl,HO)+Gl, S03. Basic Sulphate of Glycocoll. (c.) 3(G1, HO)2S03, HO. A mixture of the salt (b) with the previously described one (a), doubtless gave the crystals for the following determination: 0-3076 grm. gave 0-2300 grm. sulphate of baryta, which gives in per cent, expressed, sulphuric acid 25-65 ; corresponding with the above formula. The following is more rational. (Gl, S03, HO + G1, HO) + (Gl, HO + SO, HO). This requires 25-47 parts of sulphuric acid in 100. (d.) 2(G1, HO) + S03. Another salt gave by combustion with chromate of lead,— From 0-3039 grm., 0*2872 grm. carbonic acid, and 01680 grm. water; which expressed in per cent., give carbon 25-77, hydrogen 6-01. These numbers correspond with the formula 2(C4 H4N03,HO)+SOn, which requires carbon 25-26, and hydrogen 5 26. Glycocoll and Sulphate of Oxyd of Ethyl. Gl, HO, AeO, S03. The particular circumstances of the formation of this salt, be- yond those already given, viz. a solution in hot spirits of wine, or and some of its Products of Decomposition. 21 in water to which absolute alcohol was added, are not ascertained. With chlorid of barium, 0-6470 grm. gave 0-3036 grm. sul- phate of baryta; which in per cent, give of sulphuric acid 17-27. This quantity of acid corresponds with the formula C4 H4 N03, HO+C4 H5 O, S03, which requires 17-62 per cent, of sulphuric acid. Nitrate of Glycocoll. Gl, HO-)-NOs, HO. The capability which this compound possesses of uniting with bases enveloped the earlier conceptions of the nature of glyco- coll in obscurity:—an obscurity from which the changes the nitrate of copper salt experienced upon subjection to heat, and the simple combinations with the oxyds of silver, copper and lead, did not in any degree relieve it. It was then suggested that the glycocoll played the part of the water of crystallization in the salts that were formed. From the analysis below, it will be seen that the salts were double salts, in which glycocoll with or without water, as a base, united with hydrated nitric acid, or as a salt with nitrates of metallic oxyds. Braconnot obtained this compound by direct combination of nitric acid with glycocoll prepared from isinglass. Dessaigne procured it directly from hippuric acid, employing nitric instead of hydrochloric acid for its decomposition. We prepared it by dissolving glycocoll in strong nitric acid, and setting the solution over sulphuric acid to crystallize. Oc- casionally large tabular crystals, apparently belonging to the monoclinate system, are formed. Not unfrequently, however, the salt crystallizes in needles, especially if the fluid has been warmed. They do not deliquesce upon exposure to the air. They taste and react acid. They were dried over sulphuric acid. Combus- tion with chromate of lead gave the following results:— I. 0-4509 grm. substance gave 02954 grm. carbonic acid, artd 0-1963 grm. water. II. 0-4968 grm. substance gave 0*3122 grm. carbonic acid and 0*2054 grm. water. Two analyses, according to Varrentrapp and Will's method, gave respectively 10-04 per cent, and 1064 per cent, of nitrogen. From this it is evident that this method cannot here be employ- ed :—a fact with regard to nitrates, to which attention has already been drawn by the chemists just mentioned. Failing in this, a determination was made by the quantitative method of Prof. v. Liebig. The proportions of carbonic acid to nitrogen in four tubeSj were : 17 : 9, 14: 7, 10: 5, 24:11: or, together 65: 32=2: 1. In per cent, expressed the above determinations give 4 22 Prof. E. N. Horsford on Glycocoll, i. ii. in. Carbon, - - - 17-86 1715 Hydrogen, - - - 483 4-59 . . Nitrogen, - . . . . 20-50 These correspond with the formula C4 H4 N03, HO+N05 HO, as will be seen by the annexed estimates and results of analysis. Theory. Experiment. 1749" 4 71 20-50 57-30 7 Toboo Hydrogen, .... Nitrogen, .... Oxygen, .... 4 equiv. = 24 6 •' = 6 2 " = 28 10 " = 80 17-38 432 20-29 58-01 138 100-00 Boussingault by drying the salt at 110° C. (230° F.) obtained as already noticed the anhydrous compound C4 H4 N03, N05. Oxalate of Glycocoll. Gl, 6, HO. An aqueous solution of glycocoll with oxalic acid, evaporated upon a watch glass, crystallizes in rays reminding one of a cross section of wavellite. ' If alcohol be added to a solution of glyco- coll in oxalic acid, the latter in excess, the solution becomes milky, with the separation of oxalate of glycocoll. If added in small quantities and successively, it crystallizes with the beauty that characterizes all or nearly all the compounds of this body. Dessaigne obtained the salt directly from hippuric acid by em- ploying oxalic instead of a stronger acid, to effect the decomposi- tion. It does not alter upon exposure to the air. Combustion with chromate of lead gave the following:—0*3600 grm. gave 0*4227 grm. carbonic acid, which in per cent, express- ed, gives carbon 32-02, corresponding with the formula C4H4N03,C303,HO, which requires 32*43 per cent, of carbon. Acetate of Glycocoll. Gl, HO, A, 2HO. This salt is readily prepared by dissolving glycocoll in acetic acid, and adding absolute alcohol drop by drop, "till the solution becomes turbid, and then afterward at intervals, as the crystalli- zation proceeds. The salt analyzed was prepared by adding ab- solute alcohol in excess to a concentrated solution of glycocoll in acetic acid, (the latter in excess,) by which the salt was thrown down. It was then redissolved by heat, and set aside to cool and crystallize, by which slender prismatic crystals of great beauty were obtained. On combustion with chromate of lead, 0*2981 grm. gave 0*3644 grm. carbonie acid and 0 2031 grm. water, which in per cent. and some of its Products of Decomposition. 23 expressed correspond with carbon 33-33, hydrogen 7*57. The formula C4 H4 N03, HO+C4 H3 03-f 2HO, requires of carbon 33-33 per cent, and of hydrogen 6*94 per cent. Tartrate of Glycocoll. By dissolving glycocoll in tartaric acid and adding absolute al- cohol in excess to the solution, an oily appearing liquid separates and settles to the bottom. Repeated and protracted agitation with alcohol and ether effect no change. This liquid dried upon a watch glass gave a gummy mass which was not further inves- tigated. Palmatinate of Glycocoll. By dissolving palmitinic acid and glycocoll in hot spirits of wine, and setting aside to cool, the excess of acid rises to the sur- face in the form of an oil, while the salt crystallizes in white, thin, silky, radiating scales or blades of the greatest brilliancy. The oily layer, above, which with the whole mass becomes solid, may be readily removed, and the remainder pressed in silk and dried in the air over sulphuric acid. Combustion with chromate of lead gave the following results: i. n. in. Carbon, - 51-30 51-23 50 84 Hydrogen, - 9-45 . . 9-44 With these, no formula embracing palmitinic acid and glycocoll has been found. The formula C4 H4 NOn, Cs, H31 03-f-12HO, requires 51*31 per cent, of carbon and 11*16 per cent, hydrogen, which would correspond with the carbon, but not with the hydro- gen determinations. Glycocoll and Bi-chlorid of Platinum. Gl, PtCl,, 2HO. When to a concentrated solution of glycocoll in water, a con- centrated solution of bi-chlorid of platinum is added, and then absolute alcohol drop by drop, the solution becomes turbid, and in a very short time, regular cherry-red crystals attach themselves to the sides of the vessel. Or if the concentrated aqueous solu- tion be evaporated over sulphuric acid, after a time, groups of prismatic crystals are formed. They become instantly covered with a light colored crust upon exposure to the air, manifestly with the loss of water. 0-3679 grm. substance gave 0-0872 grm. platinum. In per cent, expressed =33-03, which corresponds with the for- mula C4 H4N03, PtCl2+2HO, which requires 33-26 per cent, of platinum. 24 Prof. E. N. Horsford on Glycocoll, Glycocoll and Chlorid of Barium. Gl, Bad, 2H0. To obtain this salt, equivalents of crystallized chlorid of bari- um (= BaCl+2HO) and glycocoll were dissolved in the least quan- tity of hot water, and suffered to crystallize quietly in the cold. In a few moments the salt crystallized in groups of short prisms of extreme beauty. None were sufficiently perfect to admit of measurement. They appeared to belong to the rhombic system, of the combination x> P. P» . xPx. The addition of alcohol to the solution changed the form to that of slender flat needles. The salt is soluble in water, more so in hot than in cold, tastes bitter, gives neither acid nor alkaline reaction, does not deliquesce or change upon exposure to the air. Dried over sulphuric acid, 0*6715 grm. substance gave 03833 grm. sulphate of baryta, =55*34 per cent, of chlorid of barium, giving the formula C4.H4 N03, BaCl, 2HO, which requires 55*31 per cent, of chlorid of barium. Glycocoll and Chlorid of Potassium. Gl, KC1. This compound was prepared by dissolving glycocoll and chlo- rid of potassium in water, and evaporating over sulphuric acid. When the solution had become very concentrated, fine needle- formed crystals filled the whole mass. They deliquesce readily in the air. A single combustion with chromate of lead, gave from 0*4992 grm., 0*3055 grm. carbonic acid= 16*58 per cent, of carbon. The formula C4 H4 N03, KC1, requires 16*92 percent, of carbon. Glycocoll and Chlorid of Sodium. A concentrated solution of glycocoll and chlorid of sodium in water, gave upon addition of absolute alcohol and standing a length of time, crystals containing both of the above mentioned ingredients. A quantitative examination was not made. Glycocoll and Bi-chlorid of Tin. By dissolving glycocoll in the least quantity of water, and ad- ding bi-chlorid of tin, after a time, crystals containing both in- gredients of the solution are formed. They were not more par- ticularly examined. Glycocoll and Hydrochlorate of Berberin. Gl, Ber, HC1. This salt is obtained by adding a hot solution of hydrochlorate of berberin in spirits of wine, to a concentrated solution, in ex- and some of its Products of Decomposition. 25 cess, of glycocoll in the same menstruum. Upon cooling, the whole mass becomes solid, and consists of myriads of the most delicate needles, of a brilliant orange color and bitter taste. The salt may be washed with water, as glycocoll is therein readily sol- uble, while the salt of berberin is not. The salt dried at 100° C. [212° Fah.] and burned with chro- mate of lead, gave the following results: 0*1563 grm. substance gave 03485 grm. carbonic acid and 0 0826 grm. water, which expressed in per cent, give carbon 60-80, hydrogen 5-87. These correspond with the formula C4 H4N03+C42H18N09, HC1, which, containing berberin with the constitution given by Fleit- mann,* requires 60-21 per cent, of carbon and 5-03 per cent, of hydrogen. Glycocoll and Potash. By dissolving glycocoll in diluted caustic potash and evapora- ting to syrup consistence over a water bath, crystals in the form of long delicate needles, containing the two ingredients, are form- ed. They may be rapidly washed with spirits of wine. They deliquesce rapidly in the air, even over sulphuric acid. Dissolved in water, the salt gives a very strong alkaline reaction. It was not further examined. Glycocoll and Hydrate of Baryta. It has already been mentioned, that glycocoll rubbed with pul- verized hydrate of baryta, in a mortar, becomes almost instanta- neously semifluid. Upon diluting the solution, and setting aside, after a time crystals containing both baryta and glycocoll were deposited. The salt was not analyzed. Its composition, in all probability, corresponds with that of the oxyd of copper, silver and lead, noticed below, and there exist, doubtless, similar salts of strontia, lime and magnesia. Glycocoll and Oxyd of Copper. Gl, CuO, HO. This salt may be prepared by adding to a solution of glycocoll sulphate of copper and caustic potash—and addition of absolute alcohol,—or by dissolving hydrated oxyd of copper, with the aid of heat, in a solution of glycocoll, and adding absolute alcohol: —or lastly by boiling the anhydrous oxyd of copper, in excess, with glycocoll. If the latter be concentrated it must be filtered hot. In this case, the filtrate in a few moments is resolved into a solid mass of the most exquisite cerulean blue color. More carefully examined, it is found to consist of exceedingly delicate * Liebig's Annalen, Bd. lix, s. 166. 26 Prof. E. N. Horsford on Glycocoll, needles. The addition of absolute alcohol to the concentrated solution precipitates the whole salt; to the diluted, less perfectly. At 100° C. [212^ p.] 0-5443 grm., at the conclusion of several days, had lost 0-0438 grm. =8-04 per cent. =one atom of water. With this loss the color passed through a light green to a shade in which a lavender or violet tint is discernible. The analysis was made with the substance dried in the air over sulphuric acid. Combustion with chromate of lead gave the following re- sults :— I. 0-2030 grm. of substance gave 0-1538 grm. carbonic acid and 00912 grm. water. II. 0-2373 grm. by the method of Varrentrapp and Will, gave 0-4762 grm. platin-salammoniac. III. 0 1745 grm. gave 00592 grm. oxyd of copper. IV. 0-2871 grm. gave 0-0972 grm. oxyd of copper. Which expressed in per cent, give I. II. III. . IV. Carbon, 20-66 Hydrogen, 4-99 . . . . . . Nitrogen, . . 12-65 Oxyd of copper, . . 33-85 33-92 These give the formula C4H4N03,CuO, HO, as will be seen by comparing the theoretical and analytical results. Theory. Experiment. 20-66 Carbon, - - - \ - 4 equiv = 24 20-92 Hydrogen, ... 5 •« = 5 4-35 4-99 Nitrogen, ... I " = 14 12-20 12-65 Oxvgen, ... 4 " = 32 27-92 27-81 Oxyd of copper, I " = 397 114 7 34-61 33-89 100-00 100-00 With the loss of an atom of water, we have the salt C4H4N03,CuO, which it will be seen is precisely the composition derived from Boussingault's analysis of the salt dried at 120° C. = [248° F.] See page 9. Glycocoll and Protoxyd of Lead. Gl, PbO, HO. This salt was prepared by dissolving with the aid of heat, prot- oxyd of lead (obtained from the peroxyd by long continued heat) in a concentrated aqueous solution of glycocoll, and the addition of alcohol till it began to be turbid. In a few hours it separated in prismatic crystals that slowly increased in size for several days, particularly with successive additions of absolute alcohol. The crystals remind one of cyanid of mercury. and some of its Products of Decomposition. 27 A single combustion with chromate of lead gave from 1*3967 grm. substance, 0-6182 grm. carbonic acid, equal to 12 07 per cent, of carbon, corresponding with the formula derived from Boussingault's analysis with the addition of an atom of water, C4H4N03,PbO,HO, which requires 12-83 per cent, of carbon. Boussingault's analysis was made from the salt, dried at 120° G, [248° F.,] leaving C4 H4 N03, PbO. Glycocoll and Oxyd of Silver. Gl, AgO, HO. If oxyd of silver be added to a solution of glycocoll, it readily dissolves with the application of heat. With the addition of alco- hol the above compound crystallizes in wartform crystals, which become dark upon exposure to light. This salt was not analyzed, as Boussingault's analysis of it, dried at 110° C. [230° F.] as already noticed, gave the formula C4H4N03,AgO. There is scarcely a doubt that corresponding compounds of co- balt, nickel, manganese and iron protoxyds with glycocoll, might with nearly equal facility be prepared. These compounds are perhaps analogous to those of ammonia with copper and nickel oxyds, when the latter are dissolved in the volatile alkali. Glycocoll and Nitrate of Silver. Gl,AgO,NOs. If the filtrate from a chlorine determination of the hydrochlo- rate of glycocoll be evaporated to concentration, and set aside over sulphuric acid, in a little time tolerably regular crystals of the above salt may be obtained. It may be procured by dissolving glycocoll in nitrate of silver: or by dissolving oxyd of silver in the solution of the nitrate of glycocoll. Upon melting, it explodes with violence. When exposed to moist air it deliquesces; though it remains unchanged over sul- phuric acid. The salt dried over sulphuric acid, on combustion with chro- mate of lead:— I. 0*9300 grm. of substance gave 03550 grm. carbonic acid and 0*1880 grm. water. II. 0*7840 grm. of the same gave 02950 grm. carbonic acid and 0*1560 grm. water. III. 0*6469 grm. of the same gave 00258 grm. chlorid of silver. In per cent, expressed, 28 Prof. E. N. Horsford on Glycocoll, i. ii. »'■ Carbon, 1011 10 26 Hydrogen. 2*24 2*21 . • Silver, . . . . 49*83 giving the formula C4 H4 N03, AgO, NO,, as the annexed estimates and results of analysis will show: | Theory. Carbon, .... 4 equiv. = 24 10 16 Hydrogen, ... 4 " = 4 1-69 Nitrogen, ... 2 " =28 11-86 Oxygen, 8 " =64 26-76 Ox. silver, ... 1 " =116 236 49-53 100 00 Experiment. "10 18 2-22 49-83 Glycocoll and Nitrate of Copper. Gl, HO, CuO, NOs, CuO, HO. This salt was analyzed by Boussingault, atid may be consid- ered as a compound of hydrate of glycocoll with nitrate of copper, united to hydrate of oxyd of copper. (Gl, HO + CuO NOs) + CuO, HO. Glycocoll and Nitrate of Potash. Gl, KO, N05. This salt forms readily from a solution of glycocoll in nitrate of potash, upon the addition of absolute alcohol. No quantitative analysis of it was made. The above formula is derived from the analyses on page 9. Glycocoll and Bi-sulphate of Potash. Gl, S03 + G1, KO, S03. By dissolving bi-sulphate of potash in water and adding a solution of glycocoll, throwing the whole down with alcohol, re- dissolving by heat and setting aside to cool and crystallize, the above salt is obtained in semi-opaque prismatic crystals. A single determination from the salt dried over sulphuric acid gave from 06873 grrn. of substance 0-6200 grm. sulph. baryta. In per cent, giving sulphuric acid =3094. The formula C4 H4 N03, S03+C4 H4 N03, KO, S03, requires of sulphuric acid 30 83 per cent. Glycocoll and Bi-chromate of Potash. If glycocoll be dissolved in an aqueous solution of bi-chromate of potash, and absolute alcohol be added till the liquid becomes turbid, and the whole set aside, in a little time crystals will be formed. These, even under the liquid, in a few days become decompo- sed, with the deposition of carbon. They were not further ex- amined. and some of its Prodncts of Decomposition. 29 Glycocoll and Urate of Ammonia. Gl, U, AmO, U. When to a hot filtered solution of urate of ammonia, glycocoll is added, in a little time as the liquid cools, long semi-opaque needles shoot out from the sides of the vessel. The addition of alcohol after the first crystallization, causes the separation of a second portion. Upon dissolving in hot water equivalents of glycocoll and urate of ammonia, and cooling, a flocculent mass was thrown down, which the addition of alcohol increased, and which, when examined with the microscope, proved to consist of exceedingly minute prisms. The salt dried over sulphuric acid and burned with chromate of lead, gave from 02926 grm. substance, 0-3463 grm. carbonic acid and 0-1144 grm. water, which equal carbon 32-46, hydrogen 4*40. The formula C4 H4 NO,, C5 N2 Ha 03+NH4 O, C5 N2 H2 0„ requires carbon 32*30, hydrogen 4*61. Similar flocculent precipitates were obtained from solutions of glycocoll in both urates of potash and soda. Glycocoll and Uric Acid. The importance of finding a compound of uric acid that would readily dissolve in water, suggested the effort to combine it with glycocoll. Two atoms of glycocoll united to two of uric acid would equal three atoms of cyanate of glycocoll: C8 H8 N2 O.+C,. N4 H4 0„=3(C, H4 NO,, C3 NO), a compound that may be presumed readily to dissolve in water. All effort to this end, however, proved unsuccessful. Uric acid remained unchanged in the most concentrated solution of glyco- coll, even with the long continued application of heat. Glycocoll and Benzoic Acid. As these two bodies exist in combination in hippuric acid, it was to be presumed that a reunion might be effected. To this end, solutions of the two in spirits of wine were made and poured together. After a time the glycocoll on the one hand and the benzoic acid on the other crystallized out. The same result attended the effort to combine cinnamic acid, cane sugar and neutral phosphate of lime with glycocoll. 5 30 Prof. E. N. Horsford on Glycocoll, Products of Decomposition of Glycocoll. Action of Sulphuric Acid. As glycocoll contains the elements of fumarate of ammonia, C4 H4 N03, HO=NH4 O, C4 H03, it was conceivable that the employment of the appropriate agen- cies might effect a simple decomposition into these two mem- bers. Neither potash nor baryta, however, were found capable of expelling the ammonia. But upon dissolving in diluted sul- phuric acid, and evaporation by heat to syrup consistence, re- dissolving with water, again evaporating, and repeating this pro- cess several times, at a certain stage, not definitely ascertained, the whole crystallizes in forms of great beauty ;—which when wash- ed with alcohol and pulverized, with addition of potash yield ammonia. Some of the crystals were nearly cubic, reminding one of chlo- rid of sodium, others were rhombic with feathered margins. They taste and react acid, and do not change upon exposure to the air. Dried over sulphuric acid and analyzed, they gave the follow- ing results. I. 0*4302 grm. of sub. gave 0*2031 carb. acid and 0*2099 water. II. 0*3526 " " " 0*8062 platin-salammoniac. HI. 0*3574 " " 0-8300 platin-salammoniac. IV. 0*4958 " " 0-6090 sulphate of baryta. Expressed in per cents. I. II. III. IV. Carbon, 12*87 Hydrogen, 5*42 . . . . . . Nitrogen, . . 14*40 14*63 . . Sulph. acid, . . . . . . 41*85 These numbers give the formula: C4 H4 N03, S03, HO+NH40, S03 HO, as the juxtaposition of the per cents, of analysis and those de- rived from direct estimate will show : Carbon, Hydrogen, Nitrogen, - Oxygen, Sulphuric acid, 4 equiv = 24 10 " = 10 2 " = 28 6 " = 48 2 « = 80 190 Theory. " 12-36 ~ Exppriim-nt. j 12-87 5-26 542 14-73 14-52 25-04 25-58 4210 41-85 100-00 100 00 This result supported the view that glycocoll contained not only the elements of fumaric acid and ammonia, but contained and some of its Products of Decomposition. 31 them in such form or arrangement, as would yield to an active affinity aided by heat. It was highly probable, therefore, that galvanic action alone would be adequate to the task of decomposition. To submit this query to the test of experiment, a solution of glycocoll was exposed to the action of a galvanic battery, con- sisting of four of Bunsen's pairs. The solution was separated by a membrane from the water. Upon closing the circle and plunging the poles terminated with platinum plate, one into the solution of glycocoll and the other into the water without, an instantaneous evolution of gas bubbles, at each pole succeded. After the action had continued a short time the fluid about the negative pole gave an alkaline reaction, while that about the positive pole gave an acid reaction. Thsre could then scarcely a doubt remain, that glycocoll was a salt of which the base was ammonia and the acid a body identical in constitution with fu- maric acid. To obtain this acid, glycocoll was treated several hours, over a moderate heat with sulphuric acid, in the manner above de- scribed—repeatedly diluting and evaporating. The sulphuric acid for the most part was thrown down with oxyd of lead, and the last traces accurately removed with solution of baryta. The filtrate was then evaporated to concentration over sulphuric acid. In a few hours rhombic prisms of unexampled beauty and per- fection, of the combination, ocP, OP, crystallized from the solution. Their taste was exceedingly sour. In water they dissolved with difficulty and in ether and alcohol they were absolutely insoluble. With potash, ammonia was evolved from them. Combustion with chromate of lead, gave the following results. I. 0*1922 grm. of sub. gave 0*1048 carb. acid and 0*1008 water. II. 0*2350 " " " 0*5700 platin-salammoniac. or in per cent.: i. n. Carbon, I486 . . Hydrogen, 5-82 . . Nitrogen, . . 15-28 With these results it was found impossible to construct any formula, containing only the elements of glycocoll, that could have been derived from the action of sulphuric acid. The first action of the acid would be the abstraction of ammonia. C4H5N04-NH3=C4H204. If in excess it was conceivable that the remaining member would be subdivided, giving oxalic acid and an oxyd of the rad- ical of Dumas' alcohol and ether series. C4H304=C203+C2H30. 32 Prof E. N. Horsfad on Glycocoll, The latter uniting with sulphuric acid would give a compound that might not be thrown down by baryta and which with the ammonia, after the precipitation of oxalic acid would give— C2 H2 O, S03, NH4 O. Upon heating the body with potash—dissolving in hydrochloric acid and adding baryta, a copious precipitate followed, establish- ing the presence of sulphuric acid. The small quantity prepared, prevented a determination of the quantity of acid. This for- mula requires 13-86 p. c. of carbon, 16-36 of nitrogen, and 6-81 p. c. of hydrogen. It is not considered as established but merely as indicating ap- proximately the action of sulphuric acid. A concentrated solution of the crystals gave with baryta a crystalline precipitate that redissolved in hydrochloric acid. With chlorid of calcium, upon the addition of ammonia, a crystalline precipitate was thrown down. Want of time as well as of substance, postponed the further examination of this interesting body. Action of Chlorine* When a moderately concentrated solution of glycocoll is sub- jected to a current of chlorine gas, the latter is rapidly absorbed, and an instantaneous and copious evolution of carbonic acid succeeds. Heat and sunlight both facilitate the action. A con- venient method was found in connecting with a stream of dry chlorine gas, a Liebig's potash apparatus, filled, as far as is usual for a combustion, with a solution of glycocoll. It is only neces- sary that the rapidity of evolution equal that of absorption. At the end of the third day the process was interrupted, and the liquid evaporated to a syrup consistence. A drop of this syrup yielded, upon the addition of ammonia, a white crystalline precipitate with both chlorids of barium and calcium. Upon saturating with baryta, filtering and washing with abso- lute alcohol, it was found that but a small fraction of the glyco- coll had been oxydated. It was again returned to the potash apparatus, and exposed to a slow but uninterrupted stream of chlorine gas for a week. At the conclusion of this period there was still glycocoll unchanged. Chlorid of barium gave the precipitate from the concentrated so- lution after neutralization with ammonia. This precipitate redis- solved in water. It also redissolved in nitric acid, from which it was not thrown down by ammonia. * Mulder did not observe any action of chlorine upon glycocoll; a circumstance attributable possibly to the presence of an impurity rich in hydrogen, or to the ac- tion not being continued sufficiently long. and some of its Products of Decomposition. 33 When washed and dried the baryta compound was no longer soluble in water, not even with long continued boiling. It was however promptly dissolved in dilute hydrochloric acid. It con- tains neither chlorine nor nitrogen. The baryta salt alone was analyzed. Combustion with chromate of lead gave from— I. 0-3218 grm. of sub. 0-1544 carb. acid and 0-0547 water. II. 0-6627 " " 0*5210 sulphate of baryta. The only formula which can be derived from these determina- tions is— C3H306-fBaO. Which requires: Estimated. Fnn Carbon, .... Hydrogen, - - - -Oxygen, ... -Baryta, .... 3 equiv. = 18 3 " = 3 6 " =48 1 " = 76-6 12-36 2-05 32-99 52-60 "100 00 13( H 33-: 5M 145-6 100- The same remarks are applicable to this formula that have been made concerning the preceding. It is recorded chiefly to show that chlorine does not act upon glycocoll as upon many other bodies, by which a certain number of atoms of hydrogen are replaced by an equal number of atoms of chlorine. The action may be conceived to be the following :— C4 H4 No3 + 5HOf3Cl = C3H3Of)+NH4 Cl + Co2-r-2HCl. The same body was obtained by direct addition of a solution of permanganate of potassa to an aqueous solution of glycocoll. After boiling a length of time with nitric acid, the same pro- duct of decomposition was formed. When pulverized chlorate of potash in small quantity and at intervals is added to a solution of glycocoll in hydrochloric acid, a slow oxydation goes forward, and a product is obtained, in which, as in the cases above noticed, baryta gives apparently the same white crystalline precipitate. Action of Caustic Potash. The brilliant fire red color assumed by glycocoll when heated with caustic potash, has already been noticed. If the solution be evaporated to extreme concentration, the evolution of ammonia and hydrogen continues, until at length the mass becomes solid. When treated with hydrochloric acid, hydrocyanic acid is evolv- ed, and if iron salts be present Berlin blue is formed. When dis- solved in water the addition of chlorid of calcium is followed by an instantaneous white precipitate, which does not dissolve in acetic acid—a precipitate of oxalate of lime. 34 Prof. E. N. Horsford on Glycocoll, The decomposition may be illustrated by the following scheme:— 1 equiv. cyanogen, 1 " ammonia, 2 " oxalic acid, . 2 " carbonic oxyd, 7 " hydrogen, 2 " glycocoll, Glycocoll and Hydrosulphuric Acid. The readiness with which glycocoll enters into combination, and the interest attaching to sulphur compounds in the products of decomposition in the "organism, suggested the treatment with hydrosulphuric acid. Taurin, according to Redtenbacher* is— C4 H7NOtf S2 which he observed contained the elements of two atoms of sul- phurous acid, aldehyde, and ammonia: C4 H, NO. S2 -=C4 H3 O, NH4 O, S2 04. The union of these several ingredients he succeeded in effecting. Taurin also contains the elements of hydrated glycocoll, two atoms of water and two of sulphur: C4 H,NOe,Sa = C4 H5N04,2HO,S2. It was conceivable that by treating a solution of glycocoll with hydrosulphuric acid, and exposing the product to oxydation, a compound, consisting of glycocoll, water, and sulphur in the above relations, might be obtained. Two atoms of Cystine : 2(C6H6N04S2) contain the elements of three atoms of hydrated glycocoll, from which ammonia has been withdrawn, and in which four atoms of oxygen have been replaced by an equal number of atoms of sulphur: Cl2 H,2 N208 S4=C)2 II,, N3 012+S4 - NH3-04 = 2(C6H6N04S,). The evolution of the latter product was also not impossible. Long continued and repeated efforts, however, gave no awaited result; the hydrated glycocoll recrystallizing from the solution with neither accession nor loss. To obtain either of the above results another experiment was made. c2 N . H? N c, 0 c2 . . 0 H7 • c8 H10 N2 0 * Liebig's Annalen, Bd. lvii, s. 170. and some of its Products of Decomposition. 35 Glycocoll was dissolved in quinqui-sulphid of potassium, spirits of wine added, and the solution evaporated through several weeks, over sulphuric acid, to dryness. An efflorescence had crept up the sides of the containing ves- sel and the bottom was covered with crystals. Upon treating the whole with hot water, sulphur was separated, which was fil- tered off and the filtrate slowly evaporated to syrup consistence, from which the whole became a solid crystalline mass. To a solution of the crystals, addition of sulphuric acid caused the evolution of sulphurous acid and the separation of sulphur. Here was the usual product of exposing the quinqui-sulphid of potassium to the air—hyposulphite of potash. Upon adding bi-chlorid of platinum to a second portion, a pre- cipitate of platin-chlorid of potassium followed, without the evo- lution of sulphurous acid or separation of sulphur. There was then left in the solution— C4 H4 N03, HO, S2 02, which equals, C4H7N06S2-H2. Constitution of Glycocoll. The enquiry presses itself, where in the general subdivisions of chemistry does glycocoll belong ? Is it a base ? Is it an acid ? Or is it a salt ? The combinations into which it is capable of entering seem only to embarrass reply. The following table of the principal compounds of glycocoll that have been analyzed, and the adjoining table of correspond- ing compounds, chiefly from inorganic chemistry, will not be without interest in the determination of this question. .Is an Acid. C4H4N03, CuO(Aq) C4 H4 NO,, PbO " C4 H4 N03, AqO « As a Base S03, CuO S03, PbO S03, AgO C4 H4 N03, SO: C. H, NO,,NO. a. AgO, SO 3 AgO, NOs C4 H4 N03, SO,, H*0 C4 H4 N03, HC1, HO C4 H4 N03, O, HO 3> b. ZnO, SO,, HO BaO, HC1, HO KO, O, HO 36 Prof E. N. Horsford on Glycocoll, c. C4 H4 N03, HO, A, 2HO | NH3, A, 2HO As a Haloid Salt. C4 H4 N03, KC1 | Hg CI, KC1. As a Bi-chlorid of Platinum. + + C4 H4 N03, Ber, HCl Pt Cl2, Ber, HC1 C4 H4 N03, BaCl, (2aq) Pt Cl2, Ba CI As Ammonia. C4 H4 N03, PtCl2 (2aq) | H6 N„, PtCl2 (Gros.) or as haloid salt, NH4 CI, PtCL 4 3 2 As an Amphid Salt. C4 H4 N03, AgO, NO, C4 H4 N03l KO, N05 HgO^O^AgO^O, As of the Magnesia Series. C4 H4 N03, S03, KO, S08.....1 C4 H4N03, S03, NH4 O, S03-t-(2aq) . . . ,2 1 . . . . MgO, SO, KO, SO,+(6aq) 2 . . . . MgO, S03 NH4 O, S03+(6aq) Basic Salts. 2 CuO N°3' H° \ N°s H0 I 3 Cu0' N°5 H0 2 (C4 H4 N03) HCl, HO | 2NH., Zn CI, HO 3 (C4 H4 N03), 2S03, 2HO | 3Hg2 O, 2N05, 3HO And further: C4 H4 N03, HO, AeO, S03 | S03 HO, AeO, S03 C4 H5 N03, U NH40, U | HO, U, NH40, U In the above tables we have representatives from almost every class of compounds in the domain of chemistry, and yet they are severally typified in the combinations of glycocoll. We have it uniting with bases;—a characteristic of acids: uniting with acids ;—a characteristic of bases : uniting with amphid salts ; a characteristic of amphid salts. We have it replacing a haloid salt in a binary chlorine com- pound ; playing the part of bichlorid of platinum in another; now appearing in the magnesia series and again replacing water. It will not be out of place to trace this relationship to other bodies in chemistry a little further, if it may aid in arriving at a nearer conception of the place glycocoll should hold. and some of its Products of Decomposition. 37 The great truth that the distinguishing properties of bodies depend upon the form, volume and density of their atoms, or to use another form of expression, that with every change of these attributes of the atoms, there is a corresponding change in the dis- tinguishing properties of the masses, is every day acquiring a more profound significance. The chemical and physical differences between phosphoric, pyrophosphoric and metaphosphoric acids, are but the counter- parts of different forms, volumes and densities assumed by the same elements in the same relative proportions. The same is true of cyanic, fulminic and cyanuric acids : of oxyd of methyle and alcohol: of hydrated acetic acid and formate of oxyd of methyle. It is well known that the several members of groups of isomor- phous bodies ;—the alums for example, have many chemical and physical attributes in common. They have also common form and volume* It requires but a little expansion of the thought naturally aris- ing from the consideration of these facts, to come upon the en- quiry: are not acids as such, indebted for their distinguishing characteristic to a common peculiarity of form among their atoms; and bases to another for theirs, and salts to another for theirs ?f Let this conception be entertained for the moment, and connect with the peculiarities of glycocoll the following considerations. 1. We have already seen that glycocoll contains the elements of fumarate of ammonia; C4H4N03=NH3,C4H03. 2. It may also be regarded as cyanate of hydrated oxyd of methyl: C4 H4N03 = C2H402,C2NO, a sugar in which alcohol and carbonic acid are replaced by wood- spirit and cyanic acid. The effort to obtain methyl alcohol by slow distillation of gly- cocoll from a solution in strong potash, failed. * Mitscherlich and Kopp. t The following interesting relationship may be worth recording. C4 H3 "o3= glycocoll. I_| C4 H3 0 03 X 3= grape sugar. Pb ^-4 H3 q 03 = sugar of lead. u C4 H3 ,j 0 03 = 1 oxyd of glyceryle. C4 H3 ij O3 HO = nitrous ether. ' These bodies have similar taste. Is it dependent upon the similar arrangement of their smallest particles? One is here reminded of the large number of acids of this formula (R) -f-03. Is the sourness of this class dependent upon a common peculiarity of form ? 6 38 Prof E. N. Horsford on Glycocoll, Dumas has suggested that the compound produced when an- hydrous sulphuric acid is conducted into an atmosphere of dry ammonia, may be considered hydrated sulphite of amidogen : NH3, S03 = HO, NH2,S02. Kane on the other hand, suggests that it be regarded as hydra- ted sulphuric acid, in which one atom of oxygen is replaced by one atom of amidogen. NH3,S03 = S ^II,)H- 3. Glycocoll may be regarded as a succinate of amidogen, C4H4NO, = NH2,C4H2 03, 4. Or as malic acid in which one atom of oxygen is replaced by amidogen. CA H, 0A - atom of malic acid. C4H4N03=C4H2]°|J 5. If doubled it becomes asparagin. 2(C4H4N03)=CSH8N206. Since the atom of water in crystallized glycocoll is replaced by both acids and bases, it may not be impossible to obtain from glycocoll a body identical in elementary composition with aspar- agin. Instances of metamorphosis of this description are not infrequent in the records of chemical investigation. The conversion of styrol into metastyrol by heat* and by vibra- tory motionf in connection with machinery, is among the most remarkable. The spontaneous change of cyanate of ammonia into urea:% of aldehyd into metaldehyd :<§> of hydrated cyanic acid into cyammelide :|| and the reconversion by distillation of cyam- melide and cyanuric acid into cyanic acid: the change of alloxan into alloxanic acid: the metamorphoses of phosphoric acid by heat, and other similar phenomena by contact with alkalies, lend support to the supposition, that it may yet be possible to effect some of the metamorphoses above suggested—or some which follow. 6. It would not be more unexpected, than was the artificial preparation of urea, to make glycocoll by combining urea and grape sugar. C^H, N,_0, + C6 H6 06 = 2(C4 Hs NOJ Urea. Half atom of grape sugar. Hydrated glycocoll. 7. The circumstance that in the preparation of the bisulphate of glycocoll and ammonia, fumaric acid is driven out, leads to * Blyihe and Hoffman, Liebig's Annalen, Bd. liii, s. 311. t Sullivan, Phil. Mag., 1845. \ VVohler. § Liebig. || Liebig and Wohler. and some of its Products of Decomposition. 39 the following consideration. If upon separating the fumaric acid, it were to unite with uncombined glycocoll, there would arise a compound containing the elements of asparaginic acid. C4H4 NQ3+C4H03 = C8H5NQ6 Fumarate of glycocoll. Asparaginic acid. 8. Two atoms of glycocoll contain also the elements of allo- phanic ether* 2(C4 H4ra);)=C4H.O,C4 H, N2 0J Glycocoll. Ether. Alloplianic acid. 9. Hydrated glycocoll contains the elements of nitric ether. C4H4NO„HO=--C4HI0,NO,. Here are two bodies of the most opposite properties, one fluid and volatile, the other solid and cannot be sublimed. We have a similar instance in the two forms of chlorid of cyanogen; the volatile discovered by Gay Lussac and the solid by Serrullas. We have another in aldehyd and metaldehyd—a liquid and a solid body. 10. Two atoms of glycocoll and one of ammonia contain the elements of the creatine of Chevreul, recently analyzed by Liebig. 2(C4 H4 NO,) +NH3= C8 H,, N3 06 .---------^---------/ ■-----------v-----------' Glycocpll. Creatine. 11. The base derived, by Liebig, from creatine, by boiling with baryta water, which separates urea (as CO, and NH3), contains the elements of glycocoll and oxyd of methyl. It con- tains also the elements of the Lactamide of Pelouze. CaH,,N,0B - C2H4N202 =C6H7N04 = C4 H4 N03, Creatine. Uiea. New base. Glycocoll. C,H,0 = C6Hs04, NH2 Oxyd of methyl. Lactamide. It will no longer seem strange that a body having so many re- lationships as are here exhibited, should find its place in the established classes of chemical compounds with difficulty. The conclusion to which we have arrived in the progress of the investigation above recorded, is, that glycocoll may at the same time be an acid, a base, and a salt, since it has properties in common with each, that distinguish each from the other two. In the possession of such a variety of attributes as attach to these three classes of bodies, glycocoll is without example in chemistry. * Liebig's Annalen, Bd. lix, s. 292. 40 Prof. E. N. Horsford on Glycocoll, Constitution of Hippuric Acid. It has long been observed that in the preparation of hippuric acid, if the heat be too high or the evaporation too rapid, ben- zoic acid alone is obtained.* If it be treated with sulphuric acid and peroxyd of manganese, carbonic acid is evolved, benzoic acid crystallizes from the hot filtered solution, and in the filtrate from the crystals sulphate of ammonia is formed.f BerzeliusJ has remarked that sulphuric acid may be considered as a compound of benzoic acid and a body of this constitution:— C4H4N03. ' Pelouze had attempted to prove that hippuric acid consisted of one atom of hydrocyanic acid, one of oil of bitter almonds, and one of formic acid: Hydrocyanic acid = C2 H N . . Hydrobenzoylic acid = Cl4 H6 . . 02 Formic acid = C2 H . . 03 Hippuric acid = Cl8 H8 N 05 Fehling$ entertained the view that it consisted of benzamide and fumaric acid. Benzamide = C14 II7 N 02 Fumaric acid — C. II . . O., 4 J Hippuric acid = C18 H8 N 05 Over all this field of speculation Dessaigne's discovery has thrown the most grateful light. Hippuric acid contains benzoic acid and glycocoll. With the aid of heat and a strong acid or alkali, the two members may be separated. C13HgN06+HCH2HO = CI4HiO:t;HO-r-C4H)N03,HCl,HO. or C.JI^+KO-fHO^C.JLO,, KO + C4H4N03, HO. By treating glycocoll with caustic potash, we obtain, among other products, cyanid of potassium and ammonia. If a stronger acid be applied to the former it yields hydrocyanic acid, which with water falls apart into formic acid and ammonia: C2NK+3H0 + HC1=KC1+NH34-C3H03. * Schwartz, Annalen dor Chem. u. Phar., liv, s. 30. Erdmann. Jour, fur Practisehe Chemie, xiii. s. 422. Dumas, by treating hippuric acid with hypochlo- rous acid, Annalesde Chem. et de Phy.*., lvii, p. 327. t Berzelius Jahresbericht, 1840, s. 701. t Jahresbericht, 1836, s. 4fj2. In the Bericht for 1831, s. 240, a similar sugges- tion from a less perfect analysis of hippuric acid, is to be found. It is again re- peated in the Bericht for 1840, s. 704. § Leibig's Annalen, Bd. xxvi, s. 60. and some of its Products of Decomposition. 41 Glycocoll contains the elements of fumaric acid and ammonia. If the former alone be taken from hippuric acid there remains benzamide. CI8H„ N05 -C^HO^C, 4 H7 NO^ Hippuric acid. Fuin;iric acid. Benzamide. Physiological Relations of Glycocoll. Ure* observed that when benzoic acid is taken into the ali- mentary canal, it reappears in the urine as hippuric acid. This at the time startling announcement, has been verified in the most satisfactory manner by Garrod,f by Keller,J in the laboratory of Prof. Booth, Philadelphia, and in the Giessen laboratory. Pettenkofer§> found in the medical examination of the urine of a girl, who suffered from St. Vitus' dance, and ate nothing but apples and bread, an unusually large quantity of hippuric acid. With the return to animal food, the abnormal proportion of this ingredient diminished. The occurrence of this acid in the urine of horses and cattle, and of men who live chiefly upon vegetable food, is well known. These facts taken in connexion with the newly developed con- stitution of hippuric acid, suggest an inquiry that may not be without interest, viz :— Are glycocoll and benzoic acid, as such, a part of the tissues of the animal body ?—of the albumen, caseine, and fibrine, sup- plied to it as food ? and finally of the corresponding bodies in the seeds and juices of plants. 1. Braconnot obtained glycocoll by treating glue with sulphuric acid ; Mulder and Boussiiig.-iult by treating glue with caustic pot- ash ; and Keller obtained it by treating the tissues or the fluids of the body with benzoic acid. The group of atoms constituting glycocoll resisted more firmly the destructive action of sulphuric acid and potash with the aid of heat, than the remaining members composing glue. These were for the most part oxydated or consumed. The albumen, fibrine, and caseine received into, and secreted from the blood, no sooner become parts of the living organism, than they commence their return to the original carbonic acid, ammonia, sulphates, phosphates, water, etc., from which they were derived. They commence oxydation. This, however, does not affect all members of the complex compounds alike. In some instances glycocoll escapes oxydation, appearing in the urine * L'lnstiiut, No. 399, 279, and No. 401, 294. Journal de Pharmacy, xxviii, p. 646. t Phil. Ma2.,xx, p. 501. t Pogg. Ann., Bd. Ivi, 638. Liebig's Ann., Bd. xliii, s. 108. Central Blatt. 1344, s. 879. § Liebig's Ann., Bd. liii, s. 86. 42 Prof. E. N. Horsford on Glycocoll, as a member of hippuric acid, as it escaped oxydation with the potash and sulphuric acid. 2. Fumaric acid is present in a great variety of plants.* In asparaginef we have the elements of fumaric acid and ammonia, which with the requisite metamorphosis would become gly- cocoll. 2(NH3, C4 H03)=C8 H8 N2 06=2(C4 H4 NO,). 3. SchlieperJ by treating isinglass with chromic acid, obtained among a variety of products benzoic acid. It is found frequently in the vegetable kingdom, for example in gum-benzoin. We have then benzoic acid and glycocoll (asparagine) in the vegetable kingdom, in the tissues of the animal body, and in the form of hippuric acid in the urine. It remains to be ascertained if they be present in the animal and vegetable albumen, fibrine, and caseine. Formation of Uric Acid Concretions. Keller observed both urea and uric acid in the urine after the separation of hippuric acid, and therefrom concludes that Ure's suggestion that benzoic acid might be employed to prevent the formation of uric acid concretions, is too hasty. The following experiment may have in connection with this subject sufficient interest to justify its being recorded. The morning urine from mixed animal and vegetable diet, was evaporated over a water bath to thick syrup consistence, and tested for glycocoll. Neither in the alcoholic extract, nor in the residue, could a trace be recognized with the oxyd of copper test. Nitric acid gave a precipitate of urea, not, however, in large quantity. At ten o'clock the next evening, four grammes of glycocoll were dissolved and taken in water. No consciousness of having taken any thing unusual was felt. The next morning urine re- acted acid. Its color was the same as that of the previous day. Upon evaporation to syrup consistence, it presented a much larger quantity than before. One portion was supersaturated with con- centrated oxalic acid, accurately neutralized with carbonate of soda and extracted with spirits of wine. Another portion was supersaturated with acetate of lead and treated with hydro-sul- phuric acid. Upon evaporating to syrup the extract of the first, and the filtrate of the second, and testing both with the oxyd of copper, no trace of glycocoll was discovered. The glycocoll had * If we include malic acid, the number will be greatly increased. Buchner's Report, Bd. xxxiv, s. 368. Liebig's Ann., Bd. xxxi, s. 241; Bd. xxxviii. s. 257; Bd. Ii, s. 246. t Geiger's Mag., xxxv, 42. J Liebig's Annalen, Bd. lviii, s. I. and some of its Products of Decomposition. 43 then disappeared. In its place were urea and uric acid, both in larger proportion, as compared with the quantities of the pre- vious day. If we deduct, as has already been shown, half an atom of grape sugar from two atoms of hydrated glycocoll, we obtain urea:— C8 H„ N2 08-C6 Hfi 0.=Ca H4 N2 02. It is conceivable that glycocoll should thus divide, and that the sugar should disappear in the products of oxydation. By treating it with nitric acid, with this view, no such result was obtained. Concentration or dilution, a strong heat or a mod- erate and long continued heat, gave no urea and no oxalic acid. One of the products is noticed on page 329. Uric acid (as bibasic) may be contemplated as a cyanurate of glycocoll:— CI0H4N4O6=C6N3O3,C4H4NO3. If benzoic acid be capable of withdrawing glycocoll from uric acid, the remaining member, in the presence of water, would furnish the materials for carbonic acid and ammonia: C6 N3 03 + 9HO-6C02-r-3NH3. No great confidence can be placed in a single result of this de- scription. Still whatever worth it has, seems to support the sug- gestion of Ure. Uric acid is found where the products of de- composition are too imperfectly oxydated.* Those who suffer from uric acid concretions are principally sedentary persons. Ex- ercise by increasing the supply and furthering the action of oxy- gen, more perfectly secures the oxydation of the products of de- composition. Could a part of the products destined to consume oxygen be withdrawn, the usual supply through the lungs might be adequate to the complete oxydation of the remainder, and thus the forma- tion of uric acid concretions be rendered impossible. Glycocoll is one of these products. If united to benzoic acid, a consumer of oxygen is withdrawn from the blood. A series of experiments upon given diet, with and without benzoic acid, could not fail to solve this important question. With this investigation, and particularly with the products of decomposition of glycocoll, which so far as here recorded, may be considered rather as qualitative than quantitative, it is our pur- pose to proceed with as little delay as possible. Note.—It may not be improper to state that Baron Liebig em- ploys his own time and that of his assistants, and the appliances of his private laboratory, in great part, in labors to ascertain methods of cheap and expeditious preparation, that he may spare * Liebig's Thier Chimie, 2e. Ausg. s. 125. 44 Prof. E. N. Horsford on Glycocoll, Sfc. the time, means, and patience of the young chemists in his school. He brings to bear his vast experience in this most difficult of all chemical labor—the preparation in their purity of chemical substances. He had been employed six months in finding out a better method than that of Braconnot or Mulder, for obtaining gelatine sugar, when in the winter semestre of 1845-46, I expressed a wish that he would give me, for a change from the labors in which I had been for some months engaged, a crystalline body, whose study would increase my knowledge of organic chemistry. In compliance with this request he gave me some three ounces of exquisitely beautiful transparent prismatic crystals, whose analysis I employed myself immediately in making. He re- marked to me of the method of preparation and of some of its properties, and of much more that I could not retain, and I went to the back journals to ascertain what investigation of it had been made; at the same time making repeated analyses of the pure body, its hydrochlorate and anhydrous sulphate. The result of this labor and a review of Boussingault's anal- yses, satisfied me that the constitution of the body, combining with acids, bases, and salts, was— C4 H4 N03. When I had come to this conclusion I had not read the article in the Comptes Rendus, containing Dessaigne's discovery, and felt indebted to no one for the constitution of the body. Gerhardt's suggestion that the body was— C2H5NO2 = O4H10N2O4.£ with his annotation, or C, II5 NO^ . \, according to that of Lie- big and Gmelin, I could not reconcile with the analysis of the anhydrous sulphate ; and this also came under my eye after my opinion of the constitution had been formed. Dessaigne had the honor of having made one of the most bril- liant of recent discoveries, but he made no analysis. Laurent made an analysis in the wake of Dessaigne's announcement, con- firming the suggestion of Dessaigne, that hippuric acid was a compound of benzoic acid and gelatine sugar. The latter how- ever is not the body, glycocoll, but its hydrate. After my paper went into the hands of the conductor of Lie- big's Annalen, Mulder's paper, giving the constitution as CtHIBNaOB in Erdmann and Marchand's Journal, appeared. This however was not the body, but its hydrate doubled.