232 SCIENCE PROGRESS 



or ants. That the practicability of this determination may 

 be realised, and that subsequent comparisons may be intelligible, 

 it will be necessary to describe the structure of these spheres 

 further. Each is seated on a short stalk of intertwined hyphae. 

 These hyphae separate above, and branch repeatedly, ultimately 

 bearing cells of two kinds : those in the interior of the sphere 

 end in a branched system of oblong or oval cells ; those on the 

 outside bear less branched chains of spherical cells, while 

 intermediate hyphae have branches bearing either or both 

 forms. Viewed under a low magnification, the sphere appears 

 to be a cluster of spherical conidia only ; but on pressing it 

 under a cover glass, the hyphae radiate, and make clear the 

 arrangement of the internal oval cells. All the hyphae spring 

 from the short stalk, and are arranged side by side, with scarcely 

 any interlacing. It is probably incorrect to style these cells 

 " conidia " — they are never shed as true conidia are. One never 

 finds on the comb free conidia or a stalk from which they have 

 fallen. If the spheres are eaten — and there is scarcely any 

 doubt of this — they must be devoured as a whole. Doflein 

 makes the quaint observation that they are "just a mouthful" 

 for the termite. 



The whole resembles one of the conidial formations which 

 are included under the head of "moulds," but it differs from 

 most of these in bearing two forms of definitely arranged 

 " conidia." If both forms of " conidia " can be found in termites 

 freshly taken from the nest, it may be assumed that the spheres 

 are eaten as a normal food. Holtermann found them in all 

 forms, Doflein found them in all but the mature workers and 

 soldiers, while I have not succeeded in finding them in the 

 larvae. The workers and soldiers obtain additional food 

 (e.g. wood, dead leaves, and other fungi) during their excursions 

 from the nest — a fact which is at once evident from the difference 

 in the colour of the contents of the intestine of these on the 

 one hand and the larvae on the other. It has been suggested 

 that the larvae destined to be sexed insects are fed always on 

 the spheres, while those which become workers and soldiers 

 are transferred at some given stage to another diet ; in other 

 words, that the differentiation into workers, soldiers, and winged 

 insects is entirely due to their food. This would seem to 

 require a segregation of each form into particular chambers, 

 some of which were destitute of fungi. This has not been shown 



