then released tbe soldier, leaving him quite limp and I fancied 

 at first poisoned, though knowing no instance of Neuropterous 

 insects secreting poison. However, the other four soldiers 

 struggled till the last vestige of their champing jaws dis- 

 appeared from view in the sand-pit. What a wretched death 

 for the Siafu, worse than being engulfed in a quagmire, but 

 not worse than those of the thousands of small insects which 

 were even now being torn from their retreats in crack and 

 cranny by the moving host. 



I presently observed a second species of ant, hereinafter 

 called the " Cocktail " [Cremastogaster castanea Sm., r. tri- 

 color Gerst.], which was also immune from attack by the 

 Siafu. These Cocktail ants have long dwelt in a corner of 

 the verandah, and at first greatly plagued me by eating; 

 insects left on the table or setting-boards, until I found how 

 to turn them to good account as the preparers of bat and 

 rodent skulls — work which they do delicately and beautifully 

 in forty-eight hours when the skulls are small. These Cock- 

 tails have regular runs and never invade the inner rooms, so 

 setting-boards were kept in an inner room and there was no 

 more trouble. The Cocktails ran about where the Siafu 

 were, but when baits were laid down the Siafu did not feed 

 at the bait used by Cocktails nor the Cocktails at that eaten 

 by the Siafu. En passant, I might say that the Cocktails 

 differ from the Siafu in that they like jam, sugar and sweet- 

 stuffs generally, while the Siafu are out for blood and fresh 

 meat only. Some species of Cocktails are very fierce, but this 

 species never bites when handled, and the only reason for- 

 their not being molested by the Siafu, as far as I can see, is; 

 that they are not juicy enough to be worth molesting. Another 

 small ant was also left in peace by the Drivers. [The examples; 

 sent were Pheidole sp., and among them a single Tetramorium 

 hlochmanni For., r. continentis For.] 



Whilst fully appreciating the Siafu as a family friend 

 assisting at a sort of spring-cleaning and ridding my residence 

 of insects of all kinds, I did not desire that they should take 

 up permanent residence should they find it a land of plenty, 

 and I was also in some trepidation as to their attitude towards 

 preserved specimens and entomological collections. There- 



