292 CLAUDE FULLER. 



widely from the larval pattern, which is retained more or less 

 by the images and workers. There are exceptions, however, 

 as in the organs of the smallest soldier found (fig. 14, t). 

 When each of a pair of antennae are composed of the same 

 number of joints they are always more or less dissimilar in 

 detail, and the pair represented in part by {p-pp) are i-ecorded 

 for their conformity as regards the basal series of joints. 

 The only statement Avhich can be made and which appears to 

 hold good very often is that when there is a difference of one 

 joint, joint III of the lesser expression is as long as III + lA^ 

 of the greater ; but there are exceptions, as may be noticed 

 by comparing the pair represented by fig. 14, n-nn. One 

 would expect to find III disproportionately large where the 

 difference runs to two joints but this is not the case 

 (cf. {o-oo) and {q-qq))- On the other hand joint III is 

 a compound joint, and is made up in some cases of three or 

 four or more joints. In all antennae it is the repository of a 

 number of joint-elements. 



Unfortunately the material before me contains very few 

 immature forms of any sort, but in a series of callow soldiers 

 of the fourth instar the antennae (fig. 14, h) show the formative 

 zone and the more recently formed joints to be swollen and 

 wider at the base than at the apex — a quite opposite condition 

 to their final form. 



To summarise, the process of joint-multiplication with 

 soldiers is determined in a very erratic manner and its sub- 

 sidence is followed by a very distinct modification. Modifica- 

 tion possibly arises to a more or less marked extent in the 

 third instar, becomes pronounced in the third pei-iod of 

 quiescency and has certain changes in the fourth instar. 



Development. — The developmentof these autennte follows 

 the common process whereby two sister-joints are abjuncted 

 at a time. The youngest forms examined have organs 

 (fig. 14,/) in stage B (PI. XV). The imago antenna (a) are 

 regularly XVI-jointed, but III is slightly variable in length 



Text-fig. 14. — Psammotermes allocerus Silv. a, h. Imago. 

 c-e. "Worker. /, g, h. Developing antennae, ^-^y. Soldier. 



