INSECTS 



highly disconcerting event if it happened anywhere else 

 than in a termite colony, where it is the regular thing. 

 But the fact of its being regular with termites makes it 

 none the less disconcerting to entomologists, for it seems 

 to defy the very laws of heredity. 



There can be no doubt of the utility of a caste system 

 where the members of each caste know their places and 

 their duties, and where nobody ever thinks of starting 

 a social revolution. But we should like to know how 

 such a svstem was ever established, and how individuals 

 of a family are not only born different but are made to 

 admit it and to act accordingly. 



These are abstruse questions, and entomologists are 

 divided in opinion as to the proper answers. Some have 

 maintained that the termite castes are not distinguished 

 when the various individuals are young, but are pro- 

 duced later by differences in the feeding — in other words, 

 it is claimed the castes are made to order by the termites 

 themselves. One particular objection to this view is that 

 no one has succeeded in finding out what the miraculous 

 pabulum may be, and no one has been able to bring about 

 a structural change in any termite by controlling its diet. 

 On the other hand, it has been shown that in some species 

 there are actual differences in the young at the time of 

 hatching, and such observations establish the fact that 

 insects from eggs laid by one female can, at least, give 

 rise to offspring of two or more forms, beside those of 

 sex, and that potential differences are determined in the 

 eggs. It is most probable that in these forms no struc- 

 tural differences could be discovered at an early embryonic 

 period, and hence it may be that, where differences are 

 not perceptible at the time of hatching, the period of 

 differentiation has only been delayed to a later stage of 

 growth. It is possible that a solution to the problem of 

 the termite castes will be found when a study of the eggs 

 themselves has been made. 



We may conclude, therefore, that the structural differ- 



[ H2] 



