PSYCHE. 



[January 1S97. 



indication of tapering as in tlie setae. 

 The arclied filaments are frequently as 

 long as the setae, and in some instances 

 longer ; their length appears to be 

 governed by the same conditions which 

 determine that of the setae. In Di- 

 plosis cucu7neris Lintn. the arched 

 filaments on the dorsal side of the 

 antennae are prolonged to nearly twice 

 the length of a segment, while on the 

 other sides they are less than half 

 that length ; the setae exhibit the 

 same variation in length. In Diplosis 

 setigera Lintn. and D. pyrivora 

 there is no such dorsal elongation of 

 the setae and the same is true of the 

 arched filaments ; they are about 

 equally developed. The outlines 

 described by the arched filaments are 

 quite variable ; they may be short with 

 a broadly rounded apex or they may be 

 excessively elongated and with a quite 

 acute apex, or they may be more fanciful 

 and describe a spatulate form. All of 

 these variations may be seen in D. 

 cucumeris. 



Dr. Enzio Reuter in a paper on 

 '■ Zwei neue Cecidomyinen " * ques- 

 tions the accuracy of the description of 

 the arched filaments given by Kiefler 

 and claims that they are but thickened 

 borders of membranous lamellae or 

 processes, and in his figures so 

 represents them. If the figures are 

 accurate, it is quite safe to conclude 

 that Dr. Reuter was looking at ver}' 

 dififerent structures from those described 



by Kiefler; he has certainly failed to 

 represent the peculiar origin of the 

 arched filaments, though he has appar- 

 ently figured the pits from which they 

 arise. In a later paper M. Kiefler and 

 M. Janet t have given in detail reasons 

 why these structures cannot be regarded 

 as the thickened borders of lamellae. 

 They found it was impossible to de- 

 monstrate the presence of a membrane 

 by histological methods ; no trace of 

 a connecting line between the bases 

 of adjoining filaments could be seen ; 

 the arched filaments changed their form 

 with great facility and in one instance a 

 seta was observed to pass through a 

 loop formed by a filament, which would 

 be unlikely to occur if a membrane 

 existed. 



In the latter article referred to above, 

 M. Janet has figured in detail the pecu- 

 liar insertion of the arched filaments and 

 he presents a theory of how they are 

 developed. In short he suggests that 

 they may have arisen as hypodermic 

 lamellae and that the edges may have 

 become thickened and chiiinized while 

 the inner membranous portion gradu- 

 ally shrank from the outhning edges 

 and disappeared. He holds the opin- 

 ion that the chitinization of the thick- 

 ened edges and shrinking of the 

 membrane at the apical portions of the 

 lamellae would, in all probabilit}-, begin 

 before they had attained their full length. 

 This is perhaps as good a theory as can 

 be advanced and agrees closely with the 



* Acta Soc. pro /ai 



\ Bull, dfs 

 .S3-1S5. 



Soc. entotnol. Fr., iSg6, pp. 37, 



