38 J. D. BERNAL 



whether this layer is referred to as clay [3], or soil [4] or even the steady con- 

 centration gradient [i] for this could only be maintained in a medium weighed 

 down by solids and the only soUd that stays put under water is clay. It is a 

 separate question to ask whether the small first-formed molecules were actually 

 absorbed on the mineral particles or merely lay in their interstices and still 

 another whether these particles also assisted in catalysing reactions between 

 organic molecules. I suspect that such catalysis did occur on particles of mont- 

 morillonite, quartz and iron hydroxide, but this can only be proved by experi- 

 ment. The argument that such structures or even their elements such as silicon 

 or aluminium are not found extensively in Uving systems to-day is beside the 

 point. There was no need for primitive life to synthesize crystals of materials 

 that were amply available in its environment and when substitutes in the form 

 of polymerized organic enzymes were formed there was no need to retain the 

 original minerals. 



As long as only small molecules formed from the condensation of carbon 

 dioxide and ammonia were available, nothing in the nature of a separate organism 

 was possible, only patches of fairly constant composition in which metaboUsm 

 was maintained for various periods and which were more like cold flames than 

 organisms as we know them. All such sub vital units would be broadly similar 

 but not identical, and would have indefinite and shifting boundaries and could 

 in general merge and fuse with one another. Where two were incompatible, one 

 or both would be destroyed; where compatible, their range of biochemical 

 activities would be increased by fusion. In the long run this would lead by a 

 kind of natural selection to improved performance and uniformity within the 

 areas in which interchange could take place, which may over extensive mud flats 

 have been many square miles in extent. 



The first crucial step which enabled Ufe to get beyond this stage and to 

 emancipate itself from mineral support was the production of polymers. These 

 must not be imagined to have had in the first place the perfection of those we 

 know in Uving things. The active polymers, the proteins and nucleic acids, are 

 characterized by high degree of specific order in the arrangements of their various 

 monomers. The passive polymers silk fibroin or cellulose seem to be simple 

 regular polymers of a better quahty than those produced artificially by most 

 radical-activated polymerization processes. 



It is not to be expected that the first natural polymers would have such a high 

 degree of regularity or specificity. This, however, would not be required in the 

 first place simply to provide the colloidal properties necessary to permit the exist- 

 ence of globules or proto-cells independent of mineral support and, owing to 

 their Ughtness, able to float away and be distributed over larger areas. These 

 proto-cells, which lacked membranes, would be formed by the loose coherence 

 of neutral hydrated particles of dimensions of the order of 150-300 A. Owing to 

 long-range ionic forces, whose precise mechanism has not yet been explained, 

 such particles cohere in quite dilute solutions. They form, as Barbu & Joly [5] 

 have shown, coacervate drops when they are approximately spherical and 

 tactoids or spindle-shaped drops when the particles are markedly elongated or 

 flattened. The spontaneous formation of such coacervate drops, varying in size 



