304 Dr. T. A. Chapman's Observations cotnpleting an 



either side with its antennae and advanced its head over 

 the gland; the larva Avould then stop walking, retreat its 

 head under the metathorax, and open rather widely the 

 6th to 7th abdominal incision, to a breadth dorsally of 

 nearly half a segment width, diminishing to either side, an 

 area conspicuous by being devoid of hairs or stellate hair- 

 bases. For some seconds the larva and ant would be 

 apparently motionless, the ant no doubt receiving honey. 

 On the first approach to this position a definite drop of 

 fluid was visible over the honey-gland just as the ant 

 approached it, when it at once disappeared. After some 

 seconds the ant left the larva, and seemed to have a deal 

 of cleaning of its legs and antennae to do; the larva at 

 once protruded its head and began to walk. Shortly the 

 ant took several walks round the larva, then approached 

 and went through the same process, and this was repeated 

 half a dozen times, except on the first occasion no drop of 

 fluid was seen, but after each occasion the ant did much 

 cleaning, though there seemed no possibility of the honey 

 or anything else having messed it. Then the carrying 

 process was begun, in what precise way the ant instructed 

 tka larva was not clear, but it assumed the bunched atti- 

 tude ; this consists, as was more clearly seen on this occa- 

 sion, in the larva swelling up the meso- and meta-thorax, 

 and so depressing the prothorax forwards as to make the 

 prothoracic plate face almost ventrally, instead of dorsally. 

 The incisions, meso-metathoracic and metathorax — 1st 

 abdominal — were in this process widely opened, showing 

 their smooth areas, and the remaining abdominal segments 

 shrunk. This particular ant did not seem to quite under- 

 stand its business ; after several attempts it seized the larva 

 by the meso-meta incision and carried the larva half an 

 inch, but the process seemed uncomfortable, at least to 

 the larva, which did not fully retain the bunched attitude, 

 and the ant let it drop. The ant after an interval again 

 milked the larva, and again got the larva to bunch, but 

 failed to get hold of it properly, and finally walked off. A 

 quarter of an hour later the larva was still walking about, 

 but after a further twenty minutes had disappeared and 

 was detected in the thick of the nest amongst ants and 

 larvae. This nest has young larvae, but not any eggs to 

 be seen. 



August 14th. — One of the larvae in the sabideti nest is 

 obviously larger than it was, or than any larva just quit- 



