394 J. D. BERNAL 



long-range order. For spherical particles only short-range order is maintained 

 usually leading to a quasi-liquid arrangement of the particles. 



In this last case we find the typical spherical coacervate droplets; in the two 

 first, the parallelism ensures an anisotropic surface tension resulting in spindle- 

 shaped drops or tactoids [13], or, when the particles are extremely elongated, 

 to swollen fibres with a regular hexagonal arrangement as in muscle [2]. 



It will be seen from this brief consideration of the nature of interparticulate 

 forces that not only do they form a sequence of decreasing strength with in- 

 creasing distance, but also that they operate between particles of increasing scale. 

 It is as if each kind of particle had to be assembled by forces of a lower order 

 until it reached the size at which the higher-order forces operate. It will be 

 noticed that the largest order of magnitude referred to, the full range of the so- 

 called long-range forces, is 1000 Â or one-tenth of a micron, still less than the 

 wavelength of visible light, so that there is still a large step between their range 

 of action and the structures observable in the light microscope. Nevertheless, 

 structures formed by means of them are responsible for the macroscopic appear- 

 ances and mechanical properties of units of sub-cellular and cellular levels. Most 

 of these ultimately depend on the tensile strength of membranes or fibres and 

 the question is therefore resolvable into that of the weakest forces holding the 

 particles of such fibres together. Owing, however, to the considerable elasticity 

 of configurational or rubber-like character of most of these structures, stresses 

 are so evenly distributed that considerable tensions and pressures can be sus- 

 tained. By combining the knowledge, still very qualitative, of the interparticle 

 forces with that of the geometrical conditions it is possible to provide some 

 approximately rational classification of the different stages of aggregation occur- 

 ring in biological structures (Table 2). 



The first stage, that of aggregation of single atoms, is covered by the theories 

 of organic chemistry. It ends with the formation of small molecular groups with 

 open or cyclic chains, most of which are found again in higher associations and 

 so may be referred to as monomers. It is noticeable that, apart from the Upids, 

 there are few if any directly atomic polymers found in organisms. It would almost 

 appear that the monomer formation and polymerization processes were intrin- 

 sically separate, or at any rate corresponded to different periods in the biopoetic 

 process. 



The second stage, the formation of polymers, is still in the field of organic 

 chemistry though in a more recent part of it and its kinetic mechanism has still 

 to be elucidated. For our purpose here it would seem that relatively few types 

 of link arc involved which may be put in some order of lability, beginning with 

 the polymetaphosphate Unk, the sugar-phosphate link occurring in nucleic acids, 

 the peptide link occurring in protein, the polysaccharide link, occurring in 

 starches and celluloses, and the carbon-carbon link, occurring in lipids and 

 rubber. These may also be, as I have indicated in my first paper (p. 38), the 

 order of occurrence in biopoesis, though I would favour, for reasons given there, 

 placing the peptide hnk before the sugar phosphate. 



The nucleic acids and the proteins differ from the other natural polymers in 

 that they are intrinsically heteropolymeric ; they are made of a number of 



