Vol. 1] Bradley. — The Taxonomy of the Masarid Wasps 371 



median parts being coiled in a great loop which may at times enor- 

 mously distend the membrane of the neck. The character is a very 

 positive one, involving the entire shape and structure of the ligula, 

 and certain chitinized basal plates. These plates seem to be developed 

 for the purpose of assisting in effecting the invaginating process, and 

 there can hardly be any transition between the two conditions. This 

 ligula is the character to which Ashmead has reference in his key to 

 genera, writing, however, by a slip of some sort, "labrum extensile" 

 or "labrum not extensile." The character may be usually made out 

 under a binocular microscope without dissection, and very readily and 

 positivelj^ by dissecting out the mouth parts ; but it is a curious fact 

 that de Saussure, owing to the poorer optical instruments available 

 in his day, has described the condition wrongly in several genera. 

 These errors have stood in literature until the present time. 



Dr. Bequaert suggested to me that the number of palpal segments 

 is of very doubtful generic value in the diplopterous wasps. The last 

 joints, he says, are very apt to drop off on the emerging of the adult, 

 or later, and furthermore there are all sorts of variations within the 

 limits of a single genus, such as Odynerus. Granting that, I can not 

 believe that the case is entirely similar in the Masaridae. My dissec- 

 tions agree perfectly for the most part with the figures and descriptions 

 of de Saussure except in the case of Celonites. I can not believe that 

 apical segments could be lost without leaving indications of the fact. 

 The shape of an ultimate segment is different from that of a penulti- 

 mate. I have examined all the North American species, and find 

 within the genus Pseudomasaris no variation in the maxillary palpi 

 (which are always rudimentary), and no variation in the labial palpi 

 of the females, which are of a peculiar type, 3-segmented, but very 

 different from the palpi of the males, even when the latter are 3-seg- 

 mented. The latter vary from 1- to 3-segmented, showing progressive 

 reduction by coalescence of the segments, as is perfectly apparent from 

 an examination of the series. In this case it would be futile to use the 

 number of segments of this sex as a generic or even subgeneric char- 

 acter. Still more surprising is the similar sexual difference in both 

 the pairs of palpi in Celonites, and especially that it should have ap- 

 parently escaped detection. Analogous sexual differences may, of 

 course, exist in other genera, and within some genera there may be 

 variation in the number of segments, as, for example, in the labial 



