NOTES ON THE FUMEIDS. 91 



■61 and -62 and the 20- jointed antennfe. I have bred one specimen of 

 this from a Eiviera larva, but have not recorded the precise locality 

 (ant. tibia, pi. iv., fig. 30). 



2Iasonia crassiorella. — I am not quite sure that I have not two 

 species confounded under this name. It should be a large species, at 

 least 16mm. in expanse, and those specimens that I regard as abso- 

 lutely agreeing with the supposed type have 24 joints to the antennas 

 and a tibial spur formula of •68--70. This form I have taken at 

 Cannes and have received from Staudinger. From another German 

 source I have received specimens with 22 joints to the antennte, and 

 the British specimens I have seen (Mr. Digby's specimen and those in 

 Dr. Mason's collection) agree in this, I have also specimens bred 

 from Locarno larvfe with only 22 antennal joints and an expanse down 

 to 12mm. All these have fairly uniform tibial spurs, varying from 

 •68--70 British, from ■66--69 the typical 24-jointed form, and •69--72 

 the Locarno form. I am unable to bring other characters into line 

 with these variations, and the number of antennal joints does not vary 

 in accord with the tibial spine. It is quite possible that my extreme 

 measurements may be somewhat in error, and that -Ql-'IO is the 

 correct tibial formula for crassiorella. Certain specimens, however, of 

 the Locarno form have a decidedly broader and shorter wing, slightly 

 more rounded at the apex. In the absence of any distinct difference 

 in spur length or antennal joints (22-23) I prefer for the present to 

 leave them as an unnamed variety. I should define 21. crassiorella as 

 a large Fnmea, 13mm. -16mm. in expanse, with 21-21 antennal joints, 

 and an anterior tibial spur of •66-*72, but this range of variation does 

 not occur everywhere, some races being of 15mm. -16mm., and with 

 24 antennal joints invariably, others, as the English forms, never 

 having 24 antennal joints, but ranging from 21-23 (pi. iv., figs. 31-34). 



B.reticulatella and B. comitella have the " cellula intrusa," that is, 

 the median nervure divides into two branches within the cell. 21. 

 crassiorella and all the other species I have examined are without it. 

 Bruand describes (or at least figures) 21. crassiorella as possessing it, 

 and from an examination of a single specimen I agreed with him. 

 Having, however, some doubt, I proceeded to make a further examina- 

 tion, and proceeded with specimen after specimen without finding it, 

 which was only to be expected, since it does not possess it. I was, 

 however, rather dissatisfied with myself at having for a time fallen into 

 so serious an error, and at length I discovered the specimen I had first 

 examined. It possesses the " cellula intrusa " Avell-developed on one side, 

 very minute, if existing at all, on the other. It was having fallen by 

 accident on this aberration that led me into my mistake. It shows 

 that one should not be satisfied with examining one specimen, even 

 when the result is to confirm a supposed well-established fact. Before 

 I discovered the source of my error I thought I might have taken one 

 of the B. reticulatella var. obscurella for an example of M. crassiorella. 

 How did Bruand fall into the error — by mere carelessness, by mis- 

 taking B. reticulatella var. for 21. crassiorella, by meeting with an 

 aberration such as I did ? I do not know, but, however much I may 

 feel annoyed at being led into error so easily, I am clearly not in a 

 position to find fault with him. 



(? Bruandia) rouasti is an eastern species. The description does 

 not show whether it is a true Fumeid or, perchance, a Proutiid. If a 



