ABILITY OF TEKMIII S TO LIVE OX PURE CELLULOSE. 2QI 



tained the termites and their food, a small amount of water was 

 added every ten days or so. In this way the food was constantly 

 kept sufficiently moist to make it wholesome and yet not so 

 moist that the termites would be injured by the growth of molds 

 on their fecal pellets. 



In each part of the second experiment, 10 nymphs were placed 

 in each of ten large *i/e sputum jars, which were kept in a moist 

 chamber. Mni-uire and food were added as in the- first experi- 

 ment. Tin- experiment differs from the !ir-t chiefly in the -i/e 

 of the \e-sel which contained the termite- ami in the number \ 

 termite* which were placed in each vessel. 



In each part of the third experiment, approximately 100 

 n> mphs \\ith i'....i! uere placed in a liter Krlenmeyer tla-k. This 

 lla-k was coinie( ted by a capillary tube to a -imilar tla-k \\hich 

 containeil a little water. The air and water were changed everj 

 on n|- 711 days, but the flask which contained tin- termite- \\a- 

 ne\er opened, for sufficient food was added when the experiment 

 w.i* -Mr ted to last to date one year and longer. 



b. The Genus Rcticnlitcrn: 



Ilit- . enus contains many individuals half the colony at 

 t of the true worker caste, while Termof>\i.\ doe* not lia\ 

 .my true \\orkcr-. In each experiment, worker-, *oldier*, and 

 nymph* \\eie used. 



Three experiments were carried out, each in two parts, u-in;^ 

 ligiKn cllnln-e in part one and cotton cellulose in the f'nrm of 

 \\'liatinan lilti-r paper No. 43 in part two, just a* with Tfrnio' 

 except that in tin- third experiment four lla-k-. with approxi- 

 mately 500 termites in each, were used. 



RESULTS. 



So I.ir as it i- po**ible to determine, the-c termites on a pun 

 eclltiln-c diet have behaved in every way exactly a* the controls, 

 that ha\e been gi\ en a wood diet and kept in the same manner. 

 No difference whatever has been noticed between those that 

 received a diet of cotton cellulose (Whatman filter paper Xo. 43) 

 and those that recei\ed a diet of wood cellulose (lignocellulose). 

 The drastic treatment to which the cotton cellulose is subjected 

 during preparation seemingly, then, does not render it more 



