326 



JOHN B. SMITH. 



inclined board, and there is a trifling distortion due to that cause ; 

 but which does not in the least interfere with accuracy of rejjresen- 

 tation. 



A close relative of the Buffalo gnat was found in a minute midge 

 swarming at Anglesea, N. J., in May, and which showed a very 

 distinct advance in mouth structure. The insect is about 1.5 mm. 

 in length, and the head does not exceed .5 mm. in diameter. The 

 task of separating out the mouth parts, small even in proportion to 

 this head, was a difficult one, and a I. objective was necessary to 

 make out all details. At a, fig. 5, the subgalea and galea are shown 



Fig. 5. Mouth parts of Anglesea midge. 



from behind, the subgalea now united at base; but the joints of the 

 galea are separate, and there is no trace of pseudo-trachea. In 

 general structure there is no great departure from that of the Bufliilo 

 gnat, except that the subgalea unite at base, and are articulated to a 

 stout chitinous piece, which may be the remnant of the cardo. There 

 is a somewhat similar structure in the mosquito at the base of the 

 " labium," and the line of development is directly from a form of 

 this description to the mosquito, in which the label la also do not con- 

 tain pseudo-trachea. In this figure we also note the method in which 

 the lacinia and palpifer enter the galear cylinder, and the paired 

 character of the structure is maintained. This is important, for the 

 membranous expansion of the labium, seen in some Odouata and in 

 Polistes (fig. 2), shows no trace of a paired structure. The tendency 

 to a fragmentary condition of the joints is also worthy of attention, 

 as here we notice the beginning of what will be afterward seen, as 

 the chitinous supports of the labella in Musea. It is noticeable here, 

 too, that the maxillary parts are not central, and arise outside the 



