,vir)4 MILLER [CHAF. '.i'l 



compounds (possibly sugars) were synthesized. These compounds were probably 

 formed from condensations of the formaldehyde that was produced by the 

 electric discharge. The alanine was demonstrated to be racemic, as would be 

 expected in a system which contained no asymmetric reagents. It was shown 

 that the syntheses were not due to bacterial contamination by performing blank 

 runs in the absence of a spark, which gave no amino acids, and by autoclaving 

 the whole apparatus prior to sparking, which gave the same yields as when the 

 autoclaving was omitted. The addition of ferrous ammonium sulfate did not 

 change the results, and the substitution of No for the NH3 changed only the 

 relative yields of the compounds produced. 



This experiment has been repeated and confirmed by Abelson (1956a), 

 Pavlovskaya and Passynskii (1957) and by Heyns, Walter and Meyer (1957). 

 Abelson worked with various mixtures of H2, CH4, CO, CO2, NH3, N2, H2O and 

 O2. As long as the conditions were reducing — either H2, CH4, CO or NH3 present 

 in excess — amino acids were synthesized. The products were the same and the 

 yields as large in many mixtures as in the methane-ammonia- water case. If the 

 conditions were oxidizing, no amino acids were synthesized. These experiments 

 have confirmed that reducing atmospheres are required for the formation of 

 organic compounds in appreciable quantities. A number of the mixtures of 

 gases used by Abelson are highly unstable. The primitive atmosphere could not 

 have been composed of such mixtures, in view of the discussion in Section 2-B. 



Heyns, Walter and Meyer also performed experiments with different mixtures 

 of gases with results similar to Abelson's. These workers also used CH4, NH3, 

 H2O and H2S. They obtained ammonium thiocyanate, thiourea and thio- 

 acetamide in addition to compounds formed when H2S was absent. All of these 

 experiments indicate that reducing conditions are required for the production 

 of appreciable amounts of these carbon compounds. 



The mechanism of synthesis of the amino acids is of interest if we are to 

 extrapolate the results in these simple systems to the primitive Earth. Two 

 alternative proposals were made for the synthesis of the amino and hydroxy 

 acids in the spark discharge system. (1) Aldehydes and hydrogen cyanide react 

 in the aqueous phase of the system to give amino and hydroxy nitriles, which 

 are hydrolyzed to amino and hydroxy acids. This mechanism is essentially a 

 Strecker synthesis. (2) The amino and hydroxy acids are synthesized in the 

 gas phase from the ions and radicals produced in the electric discharge. 



It was shown that most, if not all, of the amino acids were synthesized 

 according to the first hypothesis, since the rate of production of aldehydes and 

 hydrogen cyanide by the spark and the rate of hydrolysis of the amino nitriles 

 were sufficient to account for the total yield of amino acids (Miller, 1957). 



This mechanism accounts for the fact that most of the amino acids were 

 a-amino acids, the ones which occur in proteins. The ^-alanine was not formed 

 by this mechanism, but probably by the addition of ammonia to acrylonitrile 

 (or acrylamide or acrylic acid), followed by hydrolysis to ^-alanine. Similarly, 

 the addition of hydrogen cyanide to acrylonitrile would give the succinic acid 

 on hydrolysis. 



