PROTOZOA AND OTHER ANIMALS 969 



nympha in Kalotermes huhhardi, and probably of some forms of Dine- 

 nympha in RetJculitermes. It is also true of certain very small flagellates. 



That the flagellates possess an enzyme capable of acting on the cellu- 

 lose of the ingested wood has been clearly demonstrated by a number 

 of investigators. Trager (1932, 1934) proved that Trichomonas termop- 

 sid'is produces cellulase. He maintained the flagellate in culture for sev- 

 eral years in the presence of only one species of bacteria, which was not 

 capable of fermenting cellulose or cellobiose. The addition of finely di- 

 vided cellulose to the medium was necessary, and Trichomonas did not 

 live when that was replaced by other polysaccharides. An extract of the 

 ground bodies of the flagellates, concentrated from cultures, acted on 

 cellulose. Emik (MS) obtained fairly pure concentrations of 

 Trichonympha from Zootermopsis by gravity filtration. Extracts of these 

 concentrates were able to digest certain preparations of cellulose as 

 shown by osazone tests, demonstrating crystals of glucosazone and 

 cellobiosazone. Emik concluded that two enzymes were present, derived 

 from Trichonympha: cellulase, hydrolyzing cellulose to cellobiose; and 

 cellobiase, hydrolyzing cellobiose to glucose. 



It is not difiicult to show the action of cellulase in the contents of the 

 hind-gut, and, in view of the absence of tissue-produced cellulase and 

 the virtual absence of cellulose-digesting bacteria or fungi, the Protozoa 

 must be its source. Both cellulase and cellobiase were found there by 

 Trager (1932). Cleveland et al. (1934) and Hungate (1938) so 

 identified cellulase in flagellates of Cryptocercus punctulatus and Zoo- 

 termopsis angusticollis. 



Substance stained brown or reddish brown by iodine dissolved in 

 potassium iodide, and assumed, as is customary, to be glycogen, has 

 been found in many of these xylophagous flagellates. The earliest demon- 

 stration, which was discussed critically by Cleveland (1924), was made 

 in Trichonympha agilis by Buscalioni and Comes (1910). Kirby 

 (1932b) mentioned iodine-staining granules in T. ca?7ipanula. Yama- 

 saki (1937a, 1937b) described abundant glycogen deposits in the 

 species of Trichonympha, Teratonympha, Holomastigotes, Pyrsonympha, 

 Dinenympha, Pseudotrichonympha, Holomasti gotoides , and Spiro- 

 trichonympha in Japanese termites, preparing the material by staining 

 in Ehrlich's hematoxylin and Best's carmine after fixation in 90-percent 

 alcohol. Diminution of the glycogen in T. agilis under conditions of 



