Postulates of Network Structure

An alternative postulate to that of Gemgross is that the initial formation of a gel is due to associations between chains in pairs rather than in bundles that the cross-links in gels when first set, and particularly in dilute gels, are single rather than multiple. Further lateral association between paired chains could then eventually build up crystallites. This hypothesis would account for the existence (ff gels in which no sharp X-ray diffractions are detected. 

Schematic diagram of network formed by crystallites. From Herrmann and Gemgross (1932). 

According to this view, the first stage of gelation is the reaction between two chain segments to form a dimer, symbolised as follows  

Whether the cross-links are formed in two stages, first single and then multiple, or are all formed by bundles, the chain segments between the cross-links must be fairly flexible to explain the extensibility of gelatin gels. 

---