306 Dr. T. A. Chapman's Observations cojnpleting an 



appear to hold any communication with them, as they 

 frequently do with each other when they meet. If the 

 amount of w^atching has been sufficient, it would follow 

 that the ants do not feed the larvae, but that the latter 

 forage for themselves ; there is nothing for them to find for 

 themselves but the eggs and larvae of the ants, and con- 

 sidering that when older they eat the larvae, it seems 

 tolerably certain that at this stage they feed on the young 

 brood of the ants. 



August 19th. — The largest larva is now 6 mm. long, 

 stout in proportion, and of a fine translucent flesh-colour. 

 I succeeded in establishing an observation nest of D. flora 

 some days ago, but of a very amateurish structure ; still 

 it is an observation nest. I placed in it two larvae of 

 L. arion, but they were, I feared, not sufficiently fresh. I 

 have not since been able to discover any trace of them 

 alive or dead. I hardly think either of them can possibly 

 be there, but if this be so, I feel unable to decide whether 

 the larvae were past a condition to succeed anywhere, or 

 whether the flava were inhospitable. The pacific nature 

 of D. flava in one respect struck me : they readily accepted 

 queens picked up superficially on another nest, and sup- 

 posed to be new queens of the season as yet unprovided 

 for; but further, ants from another nest added to them 

 were fraternised with, with hardly a C[uestion asked. 



August 25th. — Larvae have been growing. At present 

 there is in var. sahideti nest a large larva 6'5 mm. long, and 

 a smaller nearly 5*0 mm., and in the scabrinodis nest one 

 nearly 6*0 mm. long; they are fond of resting on the sides 

 of the compartments away from the ants. The ants run 

 over them without paying them any attention, and I have 

 not seen one milked. 



August 26th. — The larvae do not look so well, especially 

 the one in scabrinodis looks dirty and hardly so large. The 

 others are a little dirty, seem less inclined to mix with the 

 ants, i. e. are more often seen separate from the ants, by 

 themselves, in the middle of a compartment or on a 

 slope. 



August 28th. — The scabrinodis larva looks bad, decidedly 

 shrunken, little over 4*0 mm. long. The others much as at 

 last note. 



August 29th. — The scabrinodis larva is found dead, rather 

 dried and nearlv divided into two. The sabideti ones don't 

 somehow look flourishing, but are of about the same size as 



