1901.] 55 



species of HaJictus abound there or otherwise, seems to me quite 

 unquestionable. Several such cases are mentioned in Mr. Perkins's 

 paper, and my own experience and that of others (I may venture, I 

 think, to name Mr. Saunders among these) is in the main quite in 

 accordance with his. Perhaps lists of such associations between 

 species of the two genera drawn up by observers living in different 

 parts of the country might differ slightly. Here, for instance, near 

 Woking, I am sure that iS. Jongulus is associated with H. minufissimns, 

 and nearly sure that S. puncticeps goes with H. villosulus, which are 

 cases not mentioned by Mr. Perkins, and may be confined to this or 

 certain definite districts. Everywhere S.pilifrons and H. leiicozonius, 

 S. similis and H. quadrinotatus, S. reticulatus and H. p7'asinus, seem 

 to occur together or to be alike absent. The ubiquitous 6'. (/ibhits, 

 S. siibquadratus, and S. affinis probably attach themselves to a variety 

 of species, but the special host of gihbus according to all accounts is 

 H. rubicundus, and I have in all localities found S. affinis in company 

 with S. nitidiusculus, though Mr. Perkins considers that its first 

 choice is rather H. tumulorum. Mr. Perkins has found S. spinulosus 

 with H. xanthopus, and Mr. Sladen has taken the two species in the 

 saoie place though not at the same time ; but I have never met with 

 the latter species, and on the one occasion when I found spinulosus 

 abundantly it could not, I think, have been so associated. To this 

 point I will return presently. 



So far we have only dealt with the probable parasitism of 

 Sphecodes on Halictus ; but there is at least one case in which a 

 Spkecodes has been almost proved to be parasitic on an Andrena. The 

 observations of Mr. F. W. L. Sladen and of Mr. Saunders recorded in 

 Ent. Mo. Mag. xxxiii, p. 256, and xxxiv, p. 213, make it practically 

 certain that 8. rubicundus is an inquiline to A. labialis. This is a 

 very interesting case, because 8. rubicmidus differs from almost all 

 our species in that the two sexes of each year's brood appear before 

 and not after the hot season, so that the $ has no need to hibernate 

 before ovipositing. They are ready to lay their eggs about June or 

 July, just the time when the cells made by hibernating Halicti con- 

 tain not food, but full fed young, and are closed and probably 

 unassailable by the Sphecodes. Accordingly it seems that (S^. rubi- 

 cundus is absolutely driven to have recourse to some genus with an 

 economy more like its own, and such a genus it finds in Andrena. 

 The one other British species of Sphecodes which appears (both sexes) 

 about June is the rare 8. spinulosus, and I am strongly of opinion 

 that the many $ ? of this species which I found this summer at 



