1903.] 273 



for instance, in Vespa the resemblance is so great that the difference 

 is only of a specific nature. In Psithijrus and Bombiis there is greater 

 divergence, sufficient for the creation of a genus, and yet far less than 

 in the cases of the solitary bees. It is a notable fact, that in what we 

 consider the highest aculeates, viz., the species of Apis, there should 

 be nothing of the nature of aculeate iuquilines at all. 



In the other great social section, the Heterogyna, there is only 

 one species that I know of that can be considered in any way as an 

 inquiline in this sense, viz., the extraordinary Ancrgates atratulus, 

 which lives in nests with Tetranwrium ecespitum ; this is, however, in 

 both sexes, exceedingly unlike its host. 



When we turn to the solitary genera we find wider divergence 

 in structure between the hosts and iuquilines, but in most cases the 

 male armature in both, at any rate in the higher Apidce, is formed on 

 a similar general plan, even in genera which might at first sight be 

 widely separated. Thus we find very similar armatures in AntJiophora 

 and its iuquilines Melecta and Crocisa, and to a certain extent a general 

 similarity of structure, the same may be said of the genera of the 

 Dasygastrcs, Megacliile and Coelioxys ; Clialicodoma and Dioxys, Osmia, 

 Anthidiiim, Heriades, and the species of Stelis which associate with 

 them. In all these genera the anal opening of the ,^ is inferior, and 

 I am only aware of one. case, Coelioxys qiiadridentata and Anthophora 

 parietina, as recorded by Friese and also observed by Mr. Morice, 

 where a species of this section is associated as au inquiline with 

 one of the section where the opening is terminal, or vice versa. 



The lower or short tongued bees of the families Collet idee and 

 AndrenidcB, which act as hosts, are much more widely differentiated 

 from their iuquilines (except Sphecodes and Halictus, which will be 

 considered later on). Andrena, for instance, is very different struc- 

 turally from Noinada both in its mouth parts and in the form of the 

 armature and terminal segments iu the (J j as well as in its style of 

 colouration ; still the labial palpi are cylindrical in both genera, and 

 Packard say^ that the larvae of A'omada have, like those of Andrena, 

 three conspicuous spines on the upper and posterior edge of the orbit, 

 also in both genera the stinging power is very slight. 



Nomada, however, is not exclusively associated with the Andre- 

 nidce, as a few species are known to be iuquilines on species of Eucera. 

 In Colletes and Epeolus again the difference between host and inquiline 

 is very great, both in colour and structure, but here there is one 

 property in common, viz., that both sting with unusual severity for 

 their size. The seventh ventral segment also in Epeolus ^ is not 

 altogether unlike that of Colletes in general form, i. e., if the two wings 

 of the segment and the latter were united to form a single plate. 



