THE FEEDING HABIT OF TERMITK CASTES. 303 



really too few to warrant a conclusion further than that these 

 forms certainly do not lose their protozoa and the ability to eat 

 wood as early as they do in Reticulitennes, which ha- workers. 

 If the degeneration of the jaw muscles occurs in the reproductive 

 forms when and because workers supply them with a :-alivary 

 or partly digested diet, we should not expect to find degenerate 

 jaw muscles and enlarged dependent reproductive forms in those 

 genera where workers are not present, unless, of course, the 

 young n\mph> play the same r61e that worker- do in other 

 genera. 



\\ e uill now return to the soldier caste and take up the 

 quc-tion, why are adult soldiers unable to li\e by themseh 

 Simply becau-e their very large and heavily rhitini/ed mamliblc- 

 (Fig. <> will not permit them to eat wood. They, too. like the 



nd and third forms, lose the ability to e.it \v 1 but from 



a growth process rather than one of degeneration. Before their 

 mandibles grew so large, they could eat wood and could li\e b\ 

 ihem.-el\es. Could these mandibles be altered by a change in 

 diet.' I >id a diet produce them.-* \Ye cannot answer either 

 '(notion. 



s ldiers can digest wocxl, for they have pn-to/oa in their in- 

 le-iine- to do it for them, but are unable to eat it. The\ lo-e 

 the ability to eat wood without losing their protoxoa, for they 

 pn-to/oa, just as all other castes do, very early in life tn>m 

 the ani of the xylophagous members of the colony. A> \\e ha\e 

 -aid. owing to their \ ery enormous and highly -peciali/cd man- 

 dibles (Fig. 9), they cannot chew wood, but they manage to 

 ingot proctodael wood particles partially digc-tcd perhaps 

 which ha\ e pa ed through the alimentary canal of tho-e membiT- 

 in the coloiu ca|able of chewing wood. Their intestine- are 

 con>iderabl\- Miialler and they harbor a -mailer quantity of 

 pn>to/oa than workers of the same size. Sol<lier>. then, are not 

 as dependent . in one sense, on workers or permanent wood-chew- 

 ing member^ of the colony as the second and third form- are. for 

 they do not require totally predigested food, such a- the -ali\ar\ 

 -e. ret ions upon which the second and third forms probably feed 

 entirely after the final molt. 



