50 PRO0EEDINO8 OP THE ACADEMY OP [1881. 



X. Anatomy op the Alimentary Canal in the Honey-Ant. 



These questions, closely related, required answer : 



I. Are the honej-bearers a distinct caste ? 



II. How is the peculiar dilated condition of the abdomen to be 

 accounted for ? 



III. What is the condition of the digestive organs in the ab- 

 domen of the hone3^-bearer ? 



There are some field observations that have a bearing upon 

 these questions : 



1. The workers observed returning from foraging excursions 

 had largely inflated abdomens. This is an ordinary experience 

 with ants ; the workers of Foi'mica exsectoides, our movintain 

 mound-builders, for example, returning from attendance upon the 

 Aphides with their props very much swollen. The workers of 

 Melliger, however, seem to have an especial elasticity of the crop, 

 which gave the abdomens of some of the returning repletes a 

 nearly semi-rotundity. 



2. These repletes and semi-rotunds in my artificial nests adopted 

 in a measure the sedentary habits of the honey -bearers, and perched 

 upon the roofs, where they hung quite persistently. They were 

 often very sluggish, but more ready to move than the rotunds, 

 and at times showed much activit}', though not greatl}'^ disposed 

 to work. (See PI. Ill, fig. 6.) 



3. In the formicaries opened in natural site, I observed, what 

 Llave had seen from his specimens, that there were several degrees 

 in the sizes of the honey-bearers in the honej^-rooms. 



4. There was an apparent growth in the abdomens of the seden- 

 tary' workers in the artificial nests. As early as September 7th, 

 18T9, 1 made this record in m}^ note-book: "It begins to dawn 

 upon me that the worker majors become honey-bearers. Many of 

 them hang in the nests to the honey-rooms. In 'B ' nest the en- 

 tire line along the upper margin of the large room is composed of 

 this rank." Honey -bearers with abdomens distended from one-half 

 to two-thirds the full size were continually noted, and I could onlj- 

 infer that they were recruited from the number of the sedentary 

 majors. In fact it became difficult to mark the individuals in 

 whom the sedentary major ceased and the honey-bearer began. 



5. A series of experiments was attempted to solve this point. 

 Semi-rotunds or sedentary majors were separated, freely fed, and 

 their growth noted. They never exceeded the condition of about 



