304 3Iargaret Holliday. 



The next type distingiüslied is tliat of tlie tri-ocellate worker. 

 whicli is represented by individuals with variations in tlie structure 

 of the thorax. The first type, Villa, has the typical pronotum; 

 the mesonotum and sciitellum are distinctly separated, a wide snture 

 occurs where the paraptera are expected; the sternite, epimerite 

 and episternite have fused. In the metathorax the metanotum is 

 distinct, the epinotum and sternite have fused (Fig. Cc and 

 Fig*. Ea). The second type, VIII h, has a small scutellum but 

 no mesonotum, it having fused with the epinotum and sternite 

 (Fig. Eb). The third representative of the tri-ocellate workers, 

 VII, has no suture between the pro- and meso-notum nor between 

 the mesonotum and the scutellum (Fig. E c). 



The biocellate g-roup also is represented by the three ditferent 

 forms. Tlie first having a distinct pronotum and mesonotum, and a 

 large scutellum, Via (Fig. Ed), the second having a scutellum 

 but no suture between the pro- and meso-notum, Vib (Fig. Cd 

 and Fig. E e) , the third having no scutellum and no suture 

 between the pro- and meso-notum, that is, the thorax is like that 

 of III (Fig. Ef). There is also a diiference in the arrangement 

 of the ocelli in these forms (Fig. E d, e), but this diiference occurs 

 without regard to the structure of the thorax. Some had both 

 ocelli in the posterior position; sometimes, however, the right 

 posterior ocellus and sometimes the left was present along with the 

 anterior one. 



The uni-ocellate workers also show a diiference in the ocellus 

 present, all the possible differences being shown. Sometimes the 

 anterior ocellus is present, sometimes the right posterior, sometimes 

 the left. Here also the three dififerent types, in regard to the 

 structure of the thorax are seen, just as in the other forms 

 (Fig. Ed, e. f). 



The macroergate has a small scutellum, distinctly separated 

 from the mesonotum dorsally; laterally, however, the different parts 

 of the three thoracic Segments have fused except for a short 

 distance from the ventral aspect (Fig. Cc and Fig. Eg). While 

 dissecting some of the macroergates in order to determine the 

 structure of the mouth parts one specimen, otherwise normal, was 

 found possessing antennae with twelwe joints in place of the specific 

 number, eleven, and maxillary palps with six joints whereas the 

 specific number is five. 



The microergates have the same undifterentiated structure. oiily 



