THE FEEDING HABIT OF TERMITE CASTES. 3OI 



state, when workers supply them with "partly digested food," 

 thus making the eating of wood unnecessary, may we not reason- 

 ably assume that in the second and third forms it also occurs when 

 partly digested food is supplied them by workers and wood- 

 feeding is permanently given up, that is at the time of the final 

 molt or thereabouts, considerably earlier than in the first forms. 

 If this is true, we know why the second and third form young 

 adults when experiment!}' placed by themselves do not eat wood 

 .unl die; their jaw muscles have degenerated, thus making it 

 impo--ible lor them to eat wood, and they can only survive \\ hen 

 led l.y \\orker-. 



Hut what causes the jaw muscles to degenerate' Thomson 

 and Snyder 'joj think it is due to disuse, brought about when 

 the -alivary diet takes the place of the \\<><>d diet. If thi- i- 

 true, then, the second and third forms must get more salivary 

 -e< letion or partly digested food than the lir-t form, -ince they 

 lose the ability to eat wood and, of course, their proto/oa that 

 di::e-t tin- wood for them, much earlier in lite, at least tuo years 

 earlier. If it is true that salivary secretion brings about a 

 decent i.tiioii of the jaw muscles, why are the reproductive form- 

 led .1 -alivary diet? Certainly not to make the ja\v mu-clc- 

 'iierate. tor thi- is surely only a consequence of some deeper. 

 underlying reason for feeding a salivary diet t<> the repn>din tiv e 

 lorm-. In other words, if the jaw muscle- do not dcgcnci 

 except through di-use, the salivary secretion i- pcrhap- a ix < 

 -iiv and may play a vital part either in areeleratini; or (hanging 

 the COUrse "I dexelopment. On the other hand, it" the jaw 

 muscle- degenerate in the absence of a salivary diet, that is not 

 Ix.aii-e ol di-u-e, it may be that the function ol -uch a diet is 

 -implv to lake the place of a wood diet \\hich become- im- 

 possible. It thi- is true, the question, why do the jaw mu-clc- 

 dc^cncrate. i- perhaps as vital as caste production it-ell, which, 

 it a result ot food, \\otilcl perhaps be stopped, or at least held in 

 abexaiiec to -oine extent, if individuals were isolated \cr\ earl\\ 

 A leu atteni]it- have been made to get second forms early in 

 the gradual decline of wood-feeding, which occurs simultaneously 

 with a pro-re--i\e dimunition in the number ol |udto/oa. It 

 uas found that such -econd form individuals when Elated early 

 i an live bv t hem-elves longer than if allowed to remain uith 



