in 17 Osmia-species of the adimca- Group. 175 



said to be " in der Geofend des siebenten sdiedes deutlich 

 gebogen" and "die vordere Flache ist abgeplattet, die 

 hintere aber tritt stark hervor." These, as Perez points 

 out, are characters of the present species {morawitzi sec. 

 Perez). It has a flagellum bent once (not twice as difformis) 

 about the 7th joint, and the joints up to the 10th have 

 evident posterior dilatations or "saillies" creating "une serie 

 d'echancrures." These " saillies " resemble those of dif- 

 formis, but are certainly not quite so strong, and so Perez 

 tells us, giving other minute details as to points unnoticed 

 by Morawitz. 



So far all seems satisfactory, but — 



(1) Morawitz says that his species instead of visiting 

 Echium like adunca, etc., visits " exclusively " Lotus 

 corniculatus. 



Now morawitzi, Perez, undoubtedly visits Echium, and 

 Perez gives as its plants "Echium and Lotus." 



(2) There exists another species ; differing from Perez's 

 but possessing likewise the antennal characters of loti ; 

 which (like Morawitz's /o^-types) occurs in Switzerland, 

 and which seems to me to correspond even better than 

 morawitzi, Perez, to the description of loti. 



Of this species I have two <J $ taken on the Petit Saleve 

 near Geneva, while I have only found morawitzi, Perez, in 

 South France and Algeria. 



This is the insect to which belong my Figs. 11, 20, 21, 

 27, 27a, 27b. 



I think it must be rather rare, as it seems unknown to 

 Ducke, and I have seen no specimens of it except my own. 

 (There are none in Mr. Saunders's collection, apparently.) 

 The % I do not know : and Gerstsecker says that Morawitz 

 took no females of his loti, those which he supposed to be 

 such being really only csementaria. 



It differs from morawitzi, Perez, which it strongly 

 resembles, in several important points. (1) The J 6 th 

 ventral segment (Fig. 11) is totally different, almost 

 exactly like that of difformis; (2) the femora have not 

 spinosely produced apices, as in Perez's species (Fig. 19). 

 (Unfortunately Morawitz is silent as to these characters in 

 his description of loti.) (3) The antennal joints are rather 

 more transverse. (This suits loti.) 



Another, but a trifling, difference is in the colour of the 

 antennae. These are more brightly red in the Swiss 

 species. So far as it goes, that is in favour of identifying 



