302 L. K. ( I.KYKI.AXD. 



workers for a while and isolated at a later stage of development, 

 thus showing that at the time of the earliest isolation they had 

 already begun to gradually lose the ability to maintain them- 

 selves and that they lose it more quickly when they remain 

 with workers who fed them. Thus the decline in wood-feeding 

 occurs anyway regardless of the time of the isolation, although 

 more slowly seemingly when workers are not present. However, 

 this experiment does not mean much, for long before the external 

 differentiation has occurred which makes the distinction between 

 the second and first form possible, several quite noticeably 

 distinct internal microscopic differentiations had already oc- 

 curred, thus showing that the differentiation cause had its origin 

 much earlier and had been operating for sometime before the 

 attempt to arrest it was made. Perhaps the workers could 

 distinguish them and had been feeding them. To settle the 

 question in this way, one should begin the isolation at an earlier 

 stage. But these microscopic differences just referred to are 

 only distinguishable in ReticuUtermes after fixation and staining. 

 Possibly in other genera the task will be less difficult, and we 

 may be able to determine definitely what effect, if any, a salivary 

 diet has on caste production, whether the decline in wood-feeding 

 is caused by the salivary diet or whether the salivary diet has 

 to be substituted for the wood diet after the jaw muscles have 

 degenerated. 



Since the second and third form young adults cannot live by 

 themselves, it, of course, follows that they cannot start new 

 colonies in the absence of workers. Incidently, the fact that 

 they do not harbor protozoa shows beyond question, regardless 

 of the fact that they cannot eat wood, that they never start new 

 colonies in the absence of workers. This would be true just the 

 same even if the protozoa were not absolutely necessary to their 

 existence, because if new colonies were started by second and 

 third forms, these colonies would not have protozoa in any of 

 their castes. No such colonies have ever been found; conse- 

 quently workers must be present when these forms head a colony. 



It should be mentioned here that in the genus Termopsis, which 

 has no true worker caste, it appears from the observations that 

 have been made that the ability to eat wood is not lost in the 

 second and third forms. Hut the observations on this genus are 



