368 ants' nests. 



find sugar and albuminous food upon them. These two species of 

 Acacia possess so-called extrafloral nectaries, which furnish the 

 ants with sugar, and on the points of their leaves Belt corpuscles 

 rich in albumen (resembling the MiiUer corpuscles of the Cecropia), 

 which supply them with albumen. Still, a closer direct observa- 

 tion of the reception of the food by the ants is as yet wanting 

 here. The Acacias which are free from ants do not possess these 

 peculiar arrangements 



(b) Imperfect symbiosis, — Belt has ascertained that the species 

 of Pseudomyrma which inhabit Acacia thorns are fierce, warlike 

 creatures, and keep every foe at a distance from the plant, includ- 

 ing the leaf-cutting Atta^ the forest-destroyers of America. The 

 adaptation of the ant to the plant is ascertained as soon as it is 

 proved that the respective species always lives and can thrive only 

 in the corresponding plant. This has recently been sufficiently 

 demonstrated in the case of Pseudomyrma flavidula and Pseudo- 

 myrma belti With all this, however, it is not yet proved that all 

 Acacias inhabited by ants contribute anything on their part to this 

 arrangement. In fact, this is not yet proved in the case of many 

 species ; in others, it is very doubtful or improbable, because, on 

 the one hand, there are many hollow Acacia thorns without ants 

 (Mr. Wroughton has sent me such from India), and because, on 

 the other hand, many species of ants of the genera Pseudomyrma^ 

 Sima^ and Cremastogaster frequently inhabit these thorns, and 

 frequently make their nests in some other way. These hollow 

 thorns with the round aperture, which the ants make use of, and 

 which are very similar in appearance to that of the gall in Fig. i, 

 have been often depicted, and we do not, therefore, think it worth 

 while to reproduce them here. I found a thorn of Acacia fistula^ 

 which had been brought from Somali Land by Prof. C. Keller, and 

 which was inhabited by Cremastogaster Chiarinii^ Emery, divided 

 inside by pasteboard into a few small chambers. In the case of 

 Cremastogaster chiarinii, Em., C. acacia, Forel, and C. ruspolius, 

 Forel, there appears to be an adaptation of the ant to the plant. 



We must now speak of the celebrated pseudobulbs of the epi- 

 phytic plants of the genera Myrviecodia and Hydnophytum of the 

 Sunda Islands. Fig. 14 represents, in half the natural size, the 

 photographed cross section of Hyd7iophytum moniatium^ which, 



