46 [February, 



attempt lias ever been made to clean the specimens, in order that the 

 most difficult species, such as those of the genvis Spherudes, might be 

 correctly determined ; and it is astonishing that Smith, with his great 

 love for our British bees, and with the opportunity that was his, should 

 not have taken pains to improve the condition of the collection and to 

 study it more critically. 



I have lately been permitted to clean, mount and identify some of 

 the specimens that most required to be worked at, and with the excep- 

 tion of one or two very badly mutilated examples, I believe all the 

 specimens of SpJiecodes are now capable of being distinguished. 



Kirby's actual types of each species bear a sjiecial label with a 

 number corresponding to that of the species, as numbered in his 

 "Monographia." Also the varieties are distingiiished by labels bearing 

 a Greek letter, as in that work. 



In Sphecodetf (included of course under Mel'itta) six species Avere 

 recognised by Kirby. His M. (lihha (No. 7) is a $ 5^. peJlucidvs Smith 

 (= yilifrons Thonis.), while his ^ is 8. rnhicundus von Hag. His var. /3 

 is a very old and abraded $ , hardly to be identified with certainty, but 

 possibly 8. ferruginatits Schenck. It is this specimen that he thought 

 might be a ' neiiter ' (worker) , under the supposition that 8phecodes 

 had three forms as in the social bees. 



His var. y is a. 8. .stibqvadraUts Sni., as also is the fifth example, 

 not specially labelled by him. 



Of Melitta geoffroyeJla K. there are only two females. The type is 

 a most difficult insect to name, and is, I believe, a small example of 

 8. variegatus von Hag., though some might consider it to be an ab- 

 normal example of S. diinidiatus von Hag. I can almost match it with 

 a specimen of the former from a locality where the latter does not 

 occur. The second specimen is also a small variegatus. 



Melitta spheroides K. is represented only by the 9 type. It is the 

 8phecodes gihbus of authors (nee K.). 



Melitta monilicornis K., one of which is much mutilated, having 

 lost its abdomen, the other being the type, is 8. siihqtiadratuK Sm. 



Melitta jjicea K. is represented only by the type, and that is 

 headless. It is a (J of 8. gibbva of most authors (not K.). 



There are 13 examples of M. divisa K. The type is a r? of 8. similis, 

 as also are the vars. f3, y, c (one of the two examples) and three others 

 not specially labelled. Var. 8 is headless, but is probably 8. dimidi- 

 aius ; and another labelled 12 J' 8, is ajfinis von Hag. The second 



