_M \ PETCH : THE IT NCI 



■ii:i • when hungry they may eat many things which would 



under ordinary oireumstam ses be rejected (compare Moller's 



riments). He states that a solution of the question must 



ought by an investigation of the contents of the intestine, 



m which he found the oblong conidia of the spheres in 



Ljreat abundance together with numerous vegetable fragments 



He does not say whether the insects examined were workers. 



soldiers, or larvae, or if he investigated all forms. 



Doflein (13) found the spheres in the crops of all the larva? 

 and nymphs which he investigated. He says that the crop 

 was completely filled with them, there was no other food 

 present, and the cells of the spheres were quite uninjured. 

 He declares that this is true of the larva? of the workers mid 

 soldiers as well as of the larva 1 and nymphs of the male and 

 female insects. The same observer fed all these forms with 

 the spheres, after starving them for a day : the queen also ate 

 the same, but he never succeeded in feeding a worker or 

 soldier, and in these he found nothing but finely divided 

 wood. 



Doflein suggests that the spheres form the special concen- 

 trated food of all the larva?, and the permanent food of the 

 sexed insects, while the larva? of workers and soldiers at a 

 certain stage adopt other nourishment. From this he sup- 

 poses that food plays an important pan in the differentiation 

 of workers, soldiers, and sexed forms. This is of course the 

 conclusion of (irassi. who attributes the anatomical distinc- 

 tions which arise between soldiers, workers, and winged 

 forms (of Termes lucifugus) to difference^ of food. The life 

 historic- of the present species arc as yet too imperfectly 

 known to admit of much theorizing, but one objection to 

 this view arises immediately. All the combs bear fungi, and 

 difficull tn understand how any latva> could be prevented 

 from eating them, if such had hitherto formed their normal 

 food • -p< oially in view of the fact that all these termites are 

 blind Such a method <>f differentiation would seem to 

 requin bion "t the larval destined to !>'■ workers and 



