THORAX OF THE BLOW-FLY. 19 



ferred. In general the dorsal plate of the atrophied segment is about 

 equally united with the thorax and the abdomen ; but instances 

 occur in which it approximates more closely to the former ; and 

 as this brings it nearer the structure of the Hymenoptera, I have 

 illustrated it in the case of Goerius olens (PI. II. fig. 13). It will 

 be remarked that the metathoracic postscutellum, which is usually 

 developed inwardly to form the metaphragma, is here raised to 

 the surface, forming the triangular piece between the two halves 

 of the dorsal plate of the atrophied segment, which, as usual, is 

 furnished with a pair of spiracles, and is separated by a broad 

 membranous conjunctiva from the first of the true abdominal 

 series, its lateral margins being conterminous with the epimera of 

 the ventral surface of the metathorax. Precisely the same thing 

 has happened in the Hymenoptera, both petiolated and non- 

 petiolated, only that in the former the thoracico-abdominal incisure 

 being so much deeper and taking effect more on the dorsal sur- 

 face, the union of the dorsal plate of the atrophied segment with 

 the thorax becomes more striking, and therefore seems to have 

 attracted exclusive attention. 



The phenomenon is well seen in the Humble-Bee, of which I 

 have given a drawing (PL II. fig. 6), where it will be seen how 

 large a portion of the posterior surface of the thorax is occupied 

 by this plate, reducing the metathorax in the mesial line at least 

 to a mere ridge between it and the scutellum of the mesothorax, 

 with a small triangular expansion on either side, to which the 

 bases of the posterior wings are affixed. The section of the me- 

 tathorax in the mesial line is shown in fig. 5, and it will at once 

 be seen that, viewed in this light, that segment is now reduced 

 to something like conformity with the subordinate character of 

 its alary appendages. It might be expected that these organs, 

 which are (in virtue of the hooklets by which they are united with 

 the anterior pair) evidently formed to follow the movements of 

 the latter and depend on them for their motive power, w T ould 

 require little or no provision of muscular force for themselves ; 

 and accordingly we find an almost atrophied metathorax and no 

 muscles in it. 



"We now see therefore that in the two orders of the Lepi- 

 doptera and Hymenoptera the development of the segment is 

 proportioned to the development of the wings. Surely, therefore, 

 there is good a priori ground to expect that in the Diptera the 

 same rule will hold independently of the reasons to be presently 



