308 Dr. T, A. Chapman's Observations completing an 



no recognisable or organised material, unless one so accept 

 some globules apparently of fat. 



The remaining {sabideti) larva, originally the smaller, is 

 amongst the ants and brood and looks healthy, unless a 

 blackness of the segments behind the honey-gland (possibly 

 dirt) mean black-spot disease. It is about 9"0 mm. long, 

 possibly lO'O ; it does not happen to pose well for accurate 

 measurement. Happening a little later to find the larva 

 in a more convenient position, find it measures rather over 

 9'0 mm. 



September 2nd. — Observed that the larva in sabuleti 

 nest (the only one remaining) appeared to have voided 

 some " frass." It was black, small in size, but of the 

 ordinary form of lepidopterous " frass." It was rather 

 soft. When put on a slide, pressed down and examined 

 by microscope, it was seen to contain hairs, bits of cuticle 

 and other portions of ant larvae or pupae, apparently by 

 no means small ones. The larva was measured with 

 approximate accuracy as 7"8 mm, long, 2-8 high, 3-0 wide; 

 it looks clean and healthy except the " dirt " (or whatever 

 it is) on the dorsum of 9th and 10th segments. 



Amongst the ant material of the " frass " were also seen 

 two jaws of ant larvae. 



September 6th. — Examined two (one noted under dates 

 August 2nd and 4th) nests (artificial) of D.Jlava into which 

 larvae of arioji had been introduced, and found no trace 

 of them ; this, however, goes for little, as both nests were 

 defective — one without any brood and the other hardly 

 established when the larvae were introduced. The observa- 

 tion only amounts to a failure to show that arion caji reside 

 with D.flava, and in no way shows that it cannot. 



September 8th. — Happened to notice, as I had done on 

 previous occasions, that the M. sabideti placed the debris 

 of their nest on a slip of glass used as a tray on which to 

 provide them with sugar; it occurred to me, as it ought 

 to have done before, that this debris might contain faeces 

 of the L. arion larva. On examining the debris found that 

 there were some items very like the faeces already reported 

 on, viz. small dark-coloured objects about 0-6 mm. long, 

 at first glance cylindrical, but on closer observation seen 

 to consist of two nearly spherical masses closely pressed 

 together. (The original deposit had this duplex character.) 

 One of these placed on a slide and examined under a low 

 power was seen to consist largely of hairs of the ant larvae, 



