Reprinted from SC [ E; NCE; | October 1976 Volume 194, No. 4260 AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE The Viking Biological Investigation: Preliminary Results Abstract. Three different types of biological experiments on samples of martian surface material (‘‘soil’’} were conducted inside the Viking lander. In the carbon as- similation or pyrolytic release experiment, “CO, and “CO were exposed to soil in the presence of light, A small amount of gas was found to be converted into organic mate- rial. Heat treatment of a duplicate sample prevented such conversion. In the gas ex- change experiment, soil was first humidified (exposed to water vapor) for 6 sols and then wet with a complex aqueous solution of metabolites. The gas above the soil was monitored by gas chromatography. A substantial amount of O, was detected in the Jirst chromatogram taken 2.8 hours after humidification. Subsequent analyses revealed that significant increases in CO, and only small changes in Ny had also occurred. In the labeled release experiment, soil was moistened with a solution containing several “C-labeled organic compounds. A substantial evolution of radioactive gas was regis- tered, but did not occur with a duplicate heat-treated sample. Alternative chemical and biological interpretations are possible for these preliminary data. The experi- ments are still in process, and these results so far do not allow a decision regarding the existence of life on the planet Mars. We present here a preliminary prog- ress report on the Viking biological inves- tigation, through its first month. Details of the scientific concepts behind each of the experiments, as well as examples of the kinds of results that are obtained when these concepts are tested with the use of terrestrial samples, have been de- scribed (/-3). The actual flight in- strumentation and the tests to which the flight instruments were subjected have al- so been described (4). During the manufacture of the flight in- struments for the biology experiments. rigorous clean-room techniques were em- ployed to minimize airborne con- tamination (5), after which the fully as- sembled flight hardware was heated at 120° = 1.7°C for 54 hours in an atmo- sphere of dry 100 percent nitrogen prior to shipment to the Kennedy Space Cen- ter. Here the instruments were installed in the landers under clean-room condi- tions and heated once more when the en- capsulated landers were subjected to ter- minal sterilization. This time the heating regime was 112° + 1.8°C for periods suf- ficient to reduce the spacecraft biological contamination loads to acceptable limits (6). About a month after Viking | went in- to orbit around Mars, the biology in- strument was turned on briefly for the first time since launch. At this time, 39 hours before separation. selected valves within the instrument were automatically closed to prevent exhaust products from entering the instrument during the de- scent phase when the instrument was powered down. On 22 July 1976. 2 days after landing, the instrument was again turned on. With activation of both radio- activity detectors. background counts were taken in dual- and single-channel counting modes. A chromatogram was al- so taken, and the appropriate incubation cells were rotated into position to re- ceive surface samples. The sample for the biology investigation reported here was acquired in the morning of sol 8 (a Mars day is called a sol and equals 24 hours 39 minutes) from the surface at a depth of 0 to 4 cm in an area consisting chiefly of fine-grained material. The sample was introduced into the in- strument via a soil processor on top of the lander. which screened out coarse material, larger than [.5 mm; 7 cm’ of the resulting smaller-grained material was metered down into the biology in- strument. Samples for the individual biol- ogy experiments were metered and dis- tributed into the cells for subsequent use, as described below. The temper- ature of the sample was below 0°C during acquisition and delivery, and was 9°C during the period of storage in the test cells prior to the initiation of the experi- ment. The major events for the three ex- periments are outlined in Table 1. Our overall strategy called for relative- ly short incubation periods for the first sample. If these proved negative, consid- erably longer periods could be used in lat- er incubations. Table 2 shows the vari- ous incubation sequences that are pos- sible for the three experiments. The second Viking spacecraft landed at a more northerly latitude and a colder envi- ronment. After January 1977, at this site, incubation temperatures can be signifi- cantly lowered within the biology in- strument. Part of the strategy, therefore, is to incubate martian soils at these low temperatures. The first actual science data from the Table I. Major events time line for biology investigation. Mars time Earth (sols) Events during: date 1976 fom landing Pyrolytic release Gas exchange experiment Labeled release experiment 20 July 5:12 a.m. P.D.T. 22 July 2.98 28 July 8.29 8.34 8.36 8.39 Landing Initialize instrument Acquire soil Distribute soil Seal test cell Inject “CO, and 4CO; begin incubation 8.60 9.21 9.22 29 July 9.33 10.23 10.35 11.35 11.4-13.4 13.35 13.4-13.6 31 July 2 August Begin background count Add Kr, CO,, He Inject 0.5 ml of nutrient: begin incubation Analyze gas Analyze gas Analyze gas Count background Analyze gas Terminate incubation Pyrolyze; count first peak 15.33 15.8-16.3 16.24 16.35 17.0-18.0 17.23 17.35 18.35 20.31 23.49 23.59 24.09 24.52 25.32 27.1-27.3 27.4 4 August 5 August 6 August 9 August 13 August 16 August Analyze gas Count background Inject nutrient; begin incubation Inject 2.3 ml of nutrient Analyze gas Elute second peak and count Analyze gas Analyze gas Analyze gas Analyze gas Sterilize second soil sample Inject *CO, and “CO; begin incubation 27.46 28.21 28.22 29.24 30.5-32.5 32.5-32.7 18 August Analyze gas Count background Terminate incubation; pyrolyze; count first peak 33.1-33.7 34.0-36.3 35.23 36.28* 36.51* 37.53 37.64 38.14 24 August 27 August Count background Elute second peak and count Inject nutrient Purge and dry test cell Begin background count Heat cleanup Distribute soil to second test cell Begin background count Sterilize second sample Inject nutrient; begin incubation Inject nutrient Purge test cell Count background Heat cleanup *During the interval 36.28 to 36.51, power to the entire system was interrupted, according to prior arrangement. biology instrument were returned from Mars 4 weeks before this report was writ- ten. In this interval, during which the in- strument functioned nominally, all three of the experiments yielded data in- dicating that the surface material of Mars is chemically or biochemically quite ac- tive. Under normal circumstances, it would be premature to report biological experiments in progress before the data are amenable to ready interpretation. However, the unique nature of this inves- tigation impels us to make this report, and we are fully cognizant of its prelimi- nary nature (7). The carbon assimilation experiment. The pyrolytic release (PR) or carbon as- similation experiment tests the surface material of Mars for the presence of mi- croorganisms by measuring the incorpo- ration of radioactive CO, and CO into the organic fraction of a soil sample. The reasons for believing that martian life, if it exists. would be based on carbon chemistry have been summarized (8). The experiment is carried out under ac- tual martian conditions, insofar as these can be attained within the Viking space- craft, the premise being that, if there is life on Mars, it is adapted to martian con- ditions and is probably maladapted to ex- treme departures from those conditions. The experiment operates as follows: A sample of Mars, consisting of martian at- mosphere at ambient pressure and 0.25 cm’ of soil is placed within the 4-cm’* test cell of the instrument. Martian sunlight is simulated by a 6-watt high-pressure xenon lamp, filtered to remove wave- lengths shorter than 320 nm. The ra- diant energy reaching the test chamber, integrated between 335 and 1000 nm, is approximately 20 percent of the maxi- mum solar flux at Mars in this spectral in- terval, or about 8 mw cm *. The short end of the spectrum is removed to pre- vent the surface-photecatalyzed syn- thesis of organic compounds from CO that is induced by wavelengths below 300 nm (9). Except under the special con- ditions of the photochemical synthesis, these wavelengths are generally destruc- tive to organic matter. It is therefore rea- sonably certain that, if there are orga- nisms on Mars, they have devised radia- tion protective mechanisms. Laboratory tests have shown that the experiment de- tects both tight and dark fixation of CO, and ''CO by soil microbes (/0), and the instrument can be operated in either the light or dark mode on Mars. The experi- ments so far conducted were performed in the light. The option exists to inject water vapor into the incubation cham- ber, but it was not exercised in these ex- periments. Planned Incubation of first sample (13.5 sols)* Incubation of second sample (9.5 sols) Table 2. Viking biological investigation sequences. Labeled release experiment Accomplished as of 27 August Xx xX Extended incubation before conjunction (60 sols)t Possible ‘‘through-conjunction”’ incubation (100 sols)* Extended incubation after conjunction Cold incubation, post-conjunctiont “Control” incubation if necessary Gas exchange experiment Humid incubation {7 sols) Wet incubation (30 sols) Extended. wet incubation (85 sols)* Possible. wet incubation (85 sols)* “Control” incubation, if necessary mx Pyrolytic release experiment Incubation in the light. dry (5 sols) Incubation in the light, wet (5 sols) Extended, dark incubation (35 sols) Possible ‘‘through-conjunction”’ incubationt Cold incubation, post-conjunctiont ‘Control’ incubation, if necessary *Sol, one martian day (24.6 hours). +Possible only on Viking 1. x tPossible only on Viking 2. where incubation temperatures of around 266°K can be achieved. At the siart of an experiment, 20 ul of a mixture of "CO, and "CO (92: 8 by volume, total radioactivity 22 ac) is in- jected into the test cell from a reservoir. The resulting pressure increase is 2.2 mbar over ambient which, at the Viking 1 landing site, is 7.6 mbar. The martian atmosphere is about 95 percent CO, and about 0.1 percent CO. The addition of the radioactive gases increases the par- tial pressure of CO, by 28 percent and that of CO 23-fold. The test chamber and its contents are illuminated for 120 hours at a temper- ature that depends on both the ambient martian temperature and the quantity of heat generated within the spacecraft. In the two experiments described, the in- cubation temperatures were 17° = °C and 15° + 1°C, respectively, with a brief upward excursion in the second (control) experiment to 20°C. This temperature range is clearly above the soil surface temperature at the Viking | site, where a maximum of —S5°C has been estimated during these observations (//). At the end of the incubation period, the unreacted ''CO, and CO are vented at 120°C from the test chamber. and the soil is heated to 625°C to pyrolyze any or- ganic matter it contains. The volatile products (including unreacted ''CO, and 4CO desorbed from the walls and soil particles) are swept from the chamber by a stream of He and introduced into a col- umn of Chromosorb P coated with CuO which functions as an organic vapor trap, operating at 120°C. Organic frag- ments (larger than methane) are retained by the column, but "CO, and "CO pass through and their radioactivity is counted: this count is referred to as peak 1. The column temperature is then brought to 650°C. releasing organic com- pounds and simultaneously oxidizing them to CO, by means of the CuO con- tained in the column packing. The radio- activity of this "CO, is called peak 2: it measures organic matter synthesized from "'CO, or CO during the incubation period. The results are shown in Table 3. Ex- periment | was an active experiment, conducted as described above. Experi- ment 2 was a control in which a second portion of the same surface sample was heated to 175°C for 3 hours before the start of incubation. The high background radioactivity comes primarily from two radioisotopic thermoelectric generators that supply power to the lander. Count- ing times were sufficiently long to detect approximately 10 count/min above this background. The counts were remark- ably free of noise. except during the lat- ter part of the second experiment when some noisy segments appeared. The noise was not random since the errors were all in the same (upward) direction. These segments were edited out before the data were averaged. All the counting rates summarized in Table 3 are Poisson distributed. The ‘‘expected”’ counting rates (Table 3) are those predicted if no ''C is fixed in- to organic matter. These counts repre- sent the fraction of peak 1 retained at 120°C and eluted at 650°C. This fraction is known from laboratory tests: when peak | equals 10’ count/min, the maxi- mum fraction retained is 2 x 107%, or 15 count/min for the experiments reported. Analysis of the results shows that a small but significant formation of organic matter occurred in experiment 1. The in- hibition of this process in experiment 2 shows it to be heat labile. Until a dark control is completed, we cannot know whether the fixation is light dependent. The amount of organic carbon represent- ed by 96 — 15 = 81 count/min is equiva- lent to the reduction of 7 pmole of CO or 26 pmole of CO,. Laboratory experience based on terrestrial soils suggests that two or three times more organic matter may remain in the pyrolyzed soil as a nonvolatile tar (10). Although these preliminary findings could be attributed to biological activity, several experiments remain to be done before such an interpretation can be con- sidered likely. In particular, the effect ob- served in experiment 1 must be con- firmed in a second test, and the presence of organic matter in the martian surface must be demonstrated. Given the unusu- al conditions that prevail at the surface of Mars, the possibility of nonbiological reduction of CO or CO, cannot be ex- cluded at this time. The gas exchange experiment. The gas exchange experiment (GEX) measures compositional changes in the atmo- sphere above a soil sample upon addition of aqueous nutrient medium, and from these data it attempts to show the pres- ence of microbial activity. The results from the first 20 sols of incubation show significant changes in the composition of the experimental atmosphere. GEX activities that occurred after landing, up to the end of the first in- Table 3. Pyrolytic release counting rates and their standard errors. Counts per minute Experiment Total Background Net Expected Peak | 1 (active) 7899 + 59 478 + 0.62 7421 + 59 2 (control) 8129 + 60 480 + 0.57 7649 + Peak 2 i (active) 573 + 0.83 477 + 0.79 96+ 1.15 <15 2 (control) 500 + 0.47 485 + 1.20 15+ 1.29 <15 Table 4, Gas composition (corrected) in gas exchange test cell (humid mode). The gas chromato- graph detector data are sampled at 1-second intervals, digitized, and fitted to a skewed gaussian distribution from which peak heights were obtained. The gas in the headspace is obtained from the ratio of the sample loop volume to the total headspace volume. The cumulative gas composi- tion is corrected for sampling losses by referencing absolute changes in the krypton values for successive samples. Corrections are made for pressure sensitivity in this flight instrument caused by a partial restriction in the gas sampling system which prevents total evacuation of the sample loop to ambient pressure prior to filling (three times) from the test cell. The value for krypton is corrected for pressure as follows: nanomoles Kr = 37.77 (P.)-°8 - (V,)'U where P, is the test cell pressure in millibars and V, is the peak height in volts. The value for each gas is corrected by the ratio of the term 37.77 (P,)~°"* to the similar Kr value from a pres- sure insensitive instrument. The gas composition as stated is corrected by removal of contribu- tions from known sources (for example, trace contaminants in injected gases) and for the amount dissolved in the liquid phase. An estimate of dissolved gases is made from reported values and temperature coefficients. The effects of pH on the CO, distribution are included by estimating changes in apparent CO, levels on nutrient injections in LR (second injection) and on the wet-mode nutrient injection in gas exchange. The relationship used is (nanomoles, dissolved) _ (nanomoles, gas phase) where the L values for CO, are sols 9 and 10, 21.4; on sols 11 to 15, 28.4; on sol 16, 40.4; and on sols 17, 18, 20, 25, and 28, 68.5. (volume, liquid) (volume, gas) Gas emitted (nanomoles) after humidification (hours) on Mars date: Gas (2.78) (27.86) (52.51) (101.91) (150.74) Sol 9 Sol 10 Sol 11 Sol 13 Sol 15 Nz 7 it : 16 12 8 O, 460 610 640 630 630 co, $500 9100 8800 8900 8400 Ar* 3 2 7 3 1 Net 20 20 18 20 21 Krt 2000 2000 2000 2000 1900 *Assumed to be Ar as Aris not resolved from CO on this column. tMean value for Ne, 19.88 + 0.95 (4.80 percent); mean value for Kr, 1976 + 21.54 (1.09 percent). cubation cycle, are given in Table 1. De- scriptions of the concept governing the design of the experiment and results ob- tained have been described (/2). The first incubation cycle begins with the addition of 1 cm? (73) of packed mar- tian soil to the incubation chamber. In the process of loading the soil and seal- ing the test cell on sol 8, martian atmo- sphere was trapped within the chamber at the prevailing pressure. The mixture of Kr, CO,, and He gases (/4) and 0.57 cm? (15) of aqueous nutrient medium con- taining neon were added to the test cell. This amount of nutrient was added to the bottom of the test cell so that the soil sample was contacted by water vapor on- ly, and not by the liquid medium. Results of the analyses of the headspace gases during the humid (water vapor) mode are shown in Table 4. All results are cor- rected for the initial contributions of the original trapped martian atmosphere; the added Kr, CO,, and He gas mixture; the trace amounts of gases introduced by the nutrient injection; and losses from sam- pling the headspace gas. Calculation of the actual gas concentrations is based on their partitioning between the gas and liquid phases at the incubation temper- ature (16) (Table 4). The chromatogram shows that carbon dioxide, oxygen, nitrogen, and argon and carbon monoxide (measured as a single peak) are evolved from the soil sample when warmed to 8° to 10°C and humidi- fied. The maximum amount of nitrogen gas, 16 nmole, appears on sol 11 and de- creases to one-half of this value by sol 15. Oxygen, on the other hand, after reaching its maximum on sol 11, appears to plateau. If one assumes oxidation of ascorbic acid in the medium, the actual total amount of oxygen produced equals 725 nmole (640 released into the atmo- sphere plus the 85 nmole consumed in the oxidation of the added ascorbic acid). The maximum amount of CO, pro- duced on sol 10 is approximately 9100 nmole which decreases on sol 11 to 8800 nmole. As is indicated later, the read- sorption of CO,, even after corrections for solubility, is likely associated with basicity changes in the mixture of soil and aqueous nutrient. No conclusion on the presence of CO can be drawn be- cause of the low values of the Ar and CO peak. The values of Ne and Kr demon- Strate the consistency of the internal standards and the apparent precision for the gas analyzers. The anomalous amount of O, ac- companying the desorption of CO, repre- sents an enrichment of 18 times in the martian soil. The results suggest either that molecular oxygen is held in relative- ly large quantities in the martian soil and released upon warming in the incubation test cell or that oxygen is generated from some unstable oxidant upon warming or, more likely, upon contact with water va- por. During the entire first cycle, no Hs, NO, or CH, was detected in the head- space. The absence of hydrogen upon wetting the soil seems to preclude the presence of metallic iron in concentra- tions greater than 0.003 percent. Absorption of CO, at martian surface temperatures and desorption at the in- cubation temperature of the test cell could account for some of the desorption during the 21.23 hours that the soil was sealed in the test cell. However, the data suggest that the major desorption of the CO, occurred in the 2.78 hours immedi- ately after the humidification of the test cell. These points remain to be investi- gated in the laboratory under similar con- ditions. On sol 16, an additional 2.27 cm? of nu- trient was injected. Including the amount added earlier, the nutrient now measures 2.84 cm’, and wets the soil. The data for the wet mode are shown in Table 5. The decrease in CO, seen immediately after wetting the soil may be due to pH changes of the soil-aqueous solution mix- ture. The slow rise in CO, content of the atmosphere after this initial decrease is not readily explained. This could be the result of further changes in this pH of the wet soil, or the oxidation of some of the substrates in the medium by the oxidants postulated above. That the CO, arises as a result of biological oxidation cannot, of course, be ruled out at this time. The de- crease in oxygen can be accounted for by the additional ascorbic acid in the fresh nutrient added on sol 16. The changes observed in the N, con- tent of the incubation atmosphere are Table 5. Gas composition (corrected) in gas exchange test cell (wet mode). Gas emitted (nanomoles) after 2.3 cm? of nutrient injection (hours) on Mars date Gas (2.66) (27.31) (51.98) (100.31) (223.88) (295.21) Sol 16 Sol 17 Sol 18 Sol 20 Sol 25 Sol 28 N, -6 —§ -5 —4 2 4 0, 460 380 270 210 20 210 co, 8400 9500 10400 10800 10800 10000 Ar* —3 —3 -—4 -3 -3 -3 Net 99 150 160 160 160 170 Kri 1400 1400 1400 1400 1400 1400 *As in Table 4, +Asin Table 4. tAsin Table 4. minimal and may be explained by a num- ber of processes including sorption by the soil, or by Van Slyke reactions be- tween the a-amino acids of the medium with residual nitrites in the soil. On the other hand, a biological origin (denitrifi- cation of added nitrates in the medium) is also possible. The labeled release experiment. The labeled release (LR) experiment (/, 17) seeks to detect metabolism or growth through radiorespirometry (/8). The ra- dioactive nutrient used for the test con- sists of seven simple organic substrates (formate, glycolate, glycine, b- and L- alanine, D- and L-lactate), each present at 2.5 x 10°4M and each equally and uni- formly labeled with “C (8uc/umole). To initiate the LR experiment on Mars, 0.5 cm? of the sample was placed inside a test cell, which is connected by a tube (33 by 0.2 cm, inside diameter) to another chamber flanked with two solid- state beta detectors. The background ra- dioactivity, caused primarily by the ra- dioisotopic thermoelectric generators powering the lander, was counted for ap- proximately 24 hours prior to nutrient in- jection and found to be 490 count/min. The sample was then injected with 0.115 ml of the radioactive nutrient. This vol- ume of nutrient contains approximately 257,000 count/min, each of the 17 car- & 4 digits CUMULATIVE RADIOACTIVITY 3 po iy RADIOACTIVITY CPM 3 bei by DETECTOR TEMPERATURE “i INJECTION 4 BACKGROUND Cm i Ty 2nd INJECTION bons of the seven substrates contributing approximately 15,000 count/min (corre- sponding to 29 nmole of carbon). Ap- proximately 7 sols after the first nutrient injection (Table 1), a second nutrient in- jection was made, and incubation was continued for an additional 6 sols. After each nutrient addition, radioactive gas evolved into the headspace above the sample equilibrated with the gas volume in the detector chamber. The gas accu- mulating within the detector chamber was continuously monitored for radio- activity during the incubation period. The temperature of the detector and the head end of the test cell were also moni- tored throughout the cycle. At the end of this incubation, a cycle was conducted with a second 0.5-cm! portion of the orig- inal sample held in reserve in the lander for this purpose. This was placed in a clean test cell, sealed, and heated at 170°C for 3 hours. After the cell cooled and background had been counted for ap- proximately 20 hours, nutrient was in- jected, and the evolved radioactive gas was compared to that from the first analy- sis. Details of the nutrient, instrumenta- tion, and terrestrial assays have been de- scribed (17). Upon injection of the labeled nutrient on sol 10, a vigorous production of radio- active gas was observed in the test cell as HEAD END TEMPERATURE ; EMPERATURE °o 4 a 200 10° J 200 . q . HEAT CLEAN-UP. L 4 5 : 160 4 HEAT CLEAN-UP—* 160 PURGE | ¥ 1o* t ¥ + 120 w q 120 z 3 z " 2 | 1st INJECTION : 2 . & ‘ 2nd INJECTION = leq & 7 a = Qn #0 = ed 7 : # + 4 CUMULATIVE PURGE i {BACKGROUND RADIOACTIVITY dalle | : + 40 J P j 40 do 1 DETECTOR TEMP. : 5, f J » : ° yo? LHEAD END TEMPERATURE Jo 2 16 ° ELAPSED TIME IN SOLS FROM SOL 9 T T T 4 8 2 ELAPSED TIME IN SOLS FROM SOL 28 Fig. 1 (left). Plot of labeled release data from first analysis on Mars. Radioactivity was measured at 16-minute intervals throughout the analysis cycle, except for the first 2 hours after the first nutrient injection when readings were taken every 4 minutes. Detector and head-end temperatures were measured every 16 minutes. Fig. 2 (right). Plot of labeled release data from control analysis on Mars. Radioactivity was measured at 16- minute intervals throughout the cycle, except for the first 2 hours after each nutrient injection when readings were taken every 4 minutes. Detec- tor and head-end temperatures were measured every 16 minutes. shown in Fig. 1, where data for the entire first cycle of the experiment are present- ed. The initial course of evolution of gas resembled that displayed by micro- biologically active terrestrial soils (/7). However, the rate of evolution of radio- active gas from the martian sample slowed more rapidly than would have been expected for a terrestrial soil, and approached a plateau of approximately 10,000 count/min over background. The magnitude of the response corresponds to approximately 65 percent of one of the labeled carbons in the nutrient. These facts could be an indication that only one of the substrates may have been in- volved in the reaction. Upon addition of a second volume of labeled nutrient on sol 17, an immediate (within 10 minutes) increase in evolution of radioactive gas was followed by a rap- id decrease of radioactivity until a new plateau was reached at approximately 8000 count/min. This decline accounts for approximately one-third of the total amount of gas that had been evolved, in- cluding the spike (Fig. 2) which appears immediately after the commanded nutri- ent injection. However, after reaching plateau, the radioactivity level slowly rose over the ensuing 6 sols at an average rate of approximately 40 count/min per sol. This rate is considerably less than that observed following the first in- jection. In isolating the biology instrument against the martian diurnal temperature fluctuation (approximately 187° to 242°K) at the landing site, the thermal environ- ment shown in Fig. | was imposed upon the LR module by the instrument tem- perature control system. Thus, the head end fluctuated between 9° and 13°C, and the detector temperature cycled between 14° and 26°C. Minor, regular patterns of fluctuation in the radioactivity curve cor- relate with the temperature of the test cell. Such fluctuations were anticipated and are not indicative of instrument anomalies. Thirteen sols after the first injection, cycle 1 of the LR experiment was termi- nated. To remove the accumulated radio- active gas and dry the test cell, the detec- tor and test cell were purged with heli- um. A clean test cell was then rotated under the head end, and both detectors and head end were heated during contin- uous helium purging to minimize the re- maining radioactivity. Background was then counted for about 20 hours. The new background level after the analysis averaged 516 count/min compared to the average of 490 count/min prior to the first injection. Because of the positive response tn cycle 1, a control sequence was run in cycle 2. After the control sample was heated (as described earlier), the test cell was vented to equilibrate its headspace with the martian atmosphere. After vent- ing, the radioactivity was observed to be 1300 count/min (including the 516 count/ min background), a baseline level not ex- pected to interfere seriously with the ex- periment. After acquisition of the surface sample, nutrient was delivered to the heat-treated sample. The ensuing control data are shown in Fig. 2. Some immedi- ate release of radioactive gas, totaling ap- proximately 800 count/min above the new baseline of 1300 count/min, oc- curred. However, the released gas imme- diately began to disappear from the de- tector cell, and, within about 8 hours, the radioactivity was virtually at the baseline level of 1300 count/min. After this, a slight rise in radioactivity was observed, less than that seen in the latter part of the commanded injection phase of cycle 1. Because most terrestrial control soils sterilized by heat demonstrate an imme- diate, low-level release of radioactive gas that quickly reaches a plateau and re- Mains constant, the possibility was con- sidered that the decline in radioactivity seen in Fig. 2 resulted from a gas leak in the test cell. The data obtained during background counts prior to the control show that the 1300-count/min baseline purged down to the approximate initial $16-count/min background level. Thus, radioactive gas was responsible for the elevated baseline prior to the first in- jection. If there were a leak, a reduction in the 1300 count/min would have been observed before the injection. Discussion. The experiments de- scribed above give clear evidence of chemical reactions. The essential ques- tion is whether they are attributable to a biological system. We are unable at this time to give a clear answer to that ques- tion, partly because the planned experi- mental program is not yet completed, and partly because of the inherent diffi- culty in defining complex living orga- nisms which may have developed and evolved in an environment completely different from that of the planet Earth. An important consideration in evaluat- ing the possibility of life on Mars is the chemical analysis of carbon compounds in the martian soil. Biemann ez al. (29) re- ported that no organic compounds larger than methanol and propane, for ex- ample, were observed in the Viking 1 samples at detection limits that range from 0.1 to 50 parts per billion. The re- sults are somewhat similar to those found in an Antarctic soil (No. 542, col- lected by R. E. Cameron) that has little organic material and appears not to sup- port an active biota (20). These results, especially if reinforced by analyses at a second martian site, would tend to make biology on Mars less likely, at least in the terrestrial mode. It is difficult to compare directly the re- sults of the three biology experiments since each was conducted under differ- ent conditions. Nonetheless, it is inter- esting that the two experiments dealing directly with radioactive carbon chem- istry yielded positive responses, and both were eliminated by heat steriliza- tion of the martian sample. These results violate none of the prima facie criteria for a biological process, and show some of the most general character- istics of known organisms. The positive result of the PR experiment signifies the reduction of CO or COs, or metabolic ex- change with reduced organic com- pounds, which are exhibited by all ter- restrial organisms. On the other hand, nonbiological photoreduction of CO can also be demonstrated at shorter ultravio- let wavelengths (9), and catalytic dis- mutation of CO is also well established. In contrast, the LR experiment re- quires conversion of oxidizable sub- strates into radioactive gas. In a terrestri- al test, the collective results of a positive response in cycle | and its elimination by heat sterilization in cycle 2 would sup- port the concept that microorganisms were present in the sample. The ampli- tude of the test response is an order of magnitude above that expected from a sterile soil, and the difference between the Mars test and the control cycle ex- ceeds the 3a level, which has been cho- sen as a Criterion for a positive response (17). However, important caveats to such a conclusion are (i) the possible limita- tion of metabolism to one substrate and (ii) the lack of an exponential phase of gas evolution indicative of growth. Orga- nisms in terrestrial soils attack more than one substrate, as evidenced by the fact that the plateaus attained generally repre- sent 50 percent or more of the total label added (/7). On Mars, however, utilization of only one of the offered terrestrial sub- strates might indicate a selective metabo- lism. The abrupt change in environmen- tal conditions of the martian soil imposed by the biology instrument with respect to water and temperature, together with the relatively short time of the experiment, might readily account for lack of growth. The absence of a positive response to the second injection in cycle | similar to that seen from the first injection might be at- tributed to inhibition or death of the mi- croorganisms. Despite the suggestive character of these responses of the Mars sample, the environmental conditions on Mars are sufficiently different from those on Earth to require cautious interpretation. A high ultraviolet flux strikes the martian sur- face material, and may result in the pro- duction of highly reactive compounds ca- pable of oxidizing the labeled nutrient. However, any explanation must account for the kinetics of the reaction as well as the heat lability of such oxidants or cata- lysts at 170° to 175°C. Similarly, the ab- sorption of radioactive gas after the sec- ond injection of nutrient may be facilitat- ed by alkalinity induced in the martian soil by wetting. An absorption of CO, was also seen in the GEX upon wetting the sample. Final interpretation of the results must await the results from the investigations on the second lander, the completion of Viking 1 studies, and ground-based labo- ratory experiments. HAROLD P. KLEIN NASA Ames Research Center, Moffett Field, California 94035 NorMAN H. Horowitz Division of Biology, California Institute of Technology, Pasadena 91125 GILBERT V. LEVIN Biospherics Incorporated, Rockville, Maryland 20852 VANCE I. OYAMA NASA Ames Research Center JOSHUA LEDERBERG Department of Genetics, Stanford University, Stanford, California 94305 ALEXANDER RICH Department of Biology, Massachusetts Institute of Technology, Cambridge 02139 Jerry S. HUBBARD Department of Biology, Georgia Institute of Technology, Atlanta 30332 GeorGE L. Hoppy Department of Biology, California Institute of Technology PATRICIA A. STRAAT Biospherics Incorporated BoNnNIE J. BERDAHL GLENN C. CARLE NASA Ames Research Center FREDERICK S. BROWN TRW Systems, One Space Park, Redondo Beach, California 90278 RICHARD D. JOHNSON NASA Ames Research Center References and Notes 1. N. H. Horowitz, J. $8. Hubbard, G. L. Hobby, Tearus 16, 147 (1972). 2. G. V. Levin, ibid., p. 153. 3. V. I. Oyama, ibid., p. 167. 4, 10. 11. 12. 13. 14. iS. 20. 21. H. P. Klein, Origins Life §, 431 (1974); H. P. Klein, J. Lederberg, A. Rich, N. H. Horowitz, Vv. 1. Oyama, G. V. Levin, Nature (London) 262, 24 (1976). . J. J. McDade, in Planetary Quarantine: Prin- ciples, Methods, and Problems, L. B. Hall, Ed. (Gordon & Breach, New York, 1971), pp. 37-62. . L. B. Hall, in Foundations of Space Biology and Medicine, M. Calvin and O. G. Gazenko, Eds. (Science and Technology Information Office. NASA, Washington, D.C., 1975), vol. 1, pp. 403-430. . Additional accounts of each of the three biologi- cal experiments are being prepared by the princi- pal investigators and their coinvestigators: Nor- man H. Horowitz (pyrolytic release experi- ment), Gilbert V. Levin (labeled release experi- ment), and Vance I. Oyama (gas exchange experiment). N. H. Horowitz, Accounts Chem. Res. 9, 1 (1976). . J. S. Hubbard, J. P. Hardy, N. H. Horowitz, Proc. Natl. Acad. Sci. U.S.A. 68, 474 (1971); J. S. Hubbard, J. P. Hardy, G. E. Voecks, E. E. Golub, J. Mol. Evol. 2, 149 (1973); J. S. Hub- bard, G. E. Voecks, G. L. Hobby, J. P. Ferris, E. A. Williams, D. E. Nicodem, fbid. 5, 223 (1975). J. S. Hubbard, G. L. Hobby, N. H. Horowitz, P. J. Geiger, F. A. Morelli, App!. Microbiol. 19, 32 (1970); J. S. Hubbard, Origins Life, in press. H. Kieffer, personal communication. V. 1. Oyama, B. J. Berdahl, G. C. Carle, M. E. Lehwalt, H. S. Ginoza, Origins Life, in press: E. L. Merek and V. I. Oyama, Life Sciences Space Res. 8, 108 (1970); V. I. Oyama, E. L. Merek, M. P. Silverman, C. W. Boylen, in Pro- ceedings of the Second Lunar Science Confer- ence, A. A. Levinson, Ed. (MIT Press, Cam- bridge, Mass., 1971), vol. 2, p. 1931; V. 1. Oyama, B. J. Berdahl, C. W. Boylen, E. L. Merek, in Third Lunar Science Conference, C. Watkins, Ed. (Lunar Science Institute, Hous- ton, 1972), p. 590. Solid volume of soil delivered was estimated to be 0.465 cm*. The composition of the mixture was 5.51 per- cent Kr, 2.84 percent CO,, 91.47 percent He, 0.14 percent No, 0.035 percent O.. The GEX test cell temperatures ranged from 8.3° to 10.8°C. Nutrient volume injected into the test cell esti- mated from the quantity of Ne in the headspace above the incubating soil. Neon was added to the nutrient ampule before it was sealed. . H. L. Clever and R. Battino, in Solutions and Solubilities, M. R. J. Dack, Ed. (Wiley, New York, 1975), part 1, chap. 7; T. J. Morrison and N. B. Johnstone, J. Chem. Soc. (1954), p. 3441; S. Y. Yet and R. E. Peterson, J. Pharm. Sci. 53, 822 (1964); J. D. Cos and A. J. Head. Trans. Faraday Soc. $8, 1839 (1962); W. H. Austin, E. Lacombe, P. W. Rand, M. Chatterjee, J. Appl. Physiol. 18, 301 (1963); T. J. Morrison and F. Billett, J. Chem. Soc. (1952), p. 3819. . G. V. Levin and P. A. Straat, Origins Life. in 18. 19. press. G. V. Levin, Adv. Appl. Microbiol. §. 95 (1963). K. Biemann, J. Oro, P. Toulmin II, L. E. Orgel, A. O. Nier, D. M. Anderson, P. G. Simmonds, D. Flory, A. V. Diaz, D. R. Rush- neak, J. A. Biller, Science, 194, 72 (1976). N.H. Horowitz, R. E. Cameron, J. S. Hubbard, ibid. 176, 242 (1972). We acknowledge the effective and tireless ef- forts of the engineers who are part of the Viking Biology Flight team and without whom these experiments could not have been accomplished. R. I. Gilje is chief engineer on this team, which also includes S. Loer, C. Reichwein, G. Bow- man, and D. Buckendahl. Supported, in part, by NASA contracts NAS1-9690 (to G.V.L.), NAS1-12311 (to N.H.H.), NAS1-13422 (to 1S. H.), and by NASA grants NGR-05002308 (to N.H.H.) and NSG-7069 (to J.S.H.). We also acknowledge the profound contribution to the Viking mission and particularly to the devel- opment of life detection concepts, of our former colleague and deputy team leader, Dr. Wolf ishniac. His untimely accidental death in Ant- arctica in 1973 deprived us of his keen insight and inquiring mind at a crucial time in this study. The invaluable assistance of Dr. Richard §. Young in the preparation of this manuscript is also acknowledged. 6 September 1976 Copyright © 1976 by the American Association for the Advancement of Science