316 THE ENTOMOLOGIST. 



men with fine hair, and broad at base, not at all like the 

 specimen of A. titania; hind tibia 3| mm,, hind basitarsus 

 about 2|- mm. ; hind basitarsus broadened ; venation cannot be 

 made out. The generic position of this specimen must remain 

 wholly obscure. 



A. longcBva was based on two specimens, which, judging from 

 Heer's figures, are probably not even congeneric. The specimen 

 above described is from ffiningen, but Heer's first one, from 

 Hadoboj, must be considered the true type. 



" Osmiar 

 In 1849 Heer published Osmia antiqua from G^ningen. This 

 was a poorly preserved insect, which cannot apparently be re- 

 ferred to Osmia or any other genus with certainty. In Heer's 

 work, translated and edited by Heywood, ' The Primaeval World 

 of Switzerland' (1876), vol. ii. p. 43, I find a statement that 

 there were three species of Osmia at QEningen. In the collection 

 at Ziirich I find three species from that locality, bearing manu- 

 script names by Heer. One of these, an insect about 9^ mm. 

 long, the abdomen almost 6 mm., shows no venation, and is 

 worthless for descriptive purposes. One is a wasp. The third 

 may be described as follows : — 



Andrena (/) primdva, n. sp. $ . 



Osmia primoiva, Heer, MS. A medium-sized species, with broad 

 subgiobose abdomen, clearly a bee. Thorax small ; hind legs pre- 

 served, showing scopa ; three submarginal cells. 



The hind legs are robust, formed as in Andrena, except that the 

 broad hind femur is swollen above at base ; this condition is, how- 

 ever, distinctly approached in some species of Andrena. The tibia 

 and broad basitarsus, the latter showing much long hair along its 

 hind margin, are exactly as in Andrena. The middle basitarsus is a 

 little longer than the small tibia, and is quite broad, narrowing some- 

 what toward the base. The form of this basitarsus is rather unusual, 

 but finds a close parallel in A. hattorfiana (Fabr.). 



The venation, so far as visible, is as follows : stigma long and 

 well-developed, quite normal for Andrena ; marginal cell quite normal, 

 the apex narrowly rounded, just away from costa, as in A. morio ; 

 second submarginal cell approximately square, receiving the first 

 recurrent nervure about the middle, as in A. errans ; third sub- 

 marginal cell fully twice length of second, but about equally broad 

 on marginal, receiving the second recurrent nervure just before 

 the beginning of the last third ; third transverso-cubital with a 

 single curve, not at all angulate ; second recurrent nervure normal 

 in form. 



The relatively long third submarginal cell suggests Nomia, but 

 occurs also in Andrena, e. g. A. albicans. In the hind wing the 

 marginal, cubital, and transverso-cubital nervures are visible, entirely 

 as in Andrena. The transverso-cubital is a little oblique, the lower 

 end most basad. 



