400 0. GORDON HEWITT. 



be imbedded too long or the chitin becomes brittle again. 

 Serial sections made of recently emerged imagines before the 

 chitin has hardened give good results. Other fixing agents 

 used were Perenyi, RabPs Chromoformic, Picro-formal 

 (Bourn's solution), Glacial acetic acid, and absolute alcohol. 

 Of the various stains which were used the most satisfactory 

 were Heidenhain's Iron-haamatoxylin, Brazilin, 1 and Dela- 

 fi eld's Hsematoxylin. With the last stain perfect results 

 were obtained by overstating and differentiating with acid- 

 alcohol. 



The structure of the thoracic ganglion was studied by 

 means of reconstructions. The method employed was as 

 follows : The sections were drawn by means of the camera 

 lucida on Bristol board of a thickness proportional to the 

 magnification. They were afterwards cut out and seccotined 

 together. The resulting model was trimmed and soaked in 

 melted paraffin, taken out and dipped several times till a thin 

 coating of paraffin covered the model. This was then trimmed 

 down to the original size, all the interstices having been 

 filled by the paraffin. After a coating of graphite it was 

 electrotyped with copper. In this way a permanent model 

 was obtained. 



III. EXTEENAL STRUCTURE. 

 1. The Head Capsule. 



The head capsule of M. domestica presents great modi- 

 fications when compared with the typical insect head. Con- 

 siderable difficulty is experienced in explaining its structure 

 in the morphological terms employed in the simpler orders 

 of insects. Lowne did not lessen the difficulty in describing 

 the head of the blowfly by the invention of new terms of 

 little morphological value. The head of the fly is strongly 

 convex in front (PL 23, fig. 1), the posterior surface being 

 almost flat and slightly conical. For the sake of clearness the 



1 See Hickson, S. J., " Staining with Brazilin," ' Quart. Journ. Micr. Sci./ 

 vol. 44, pp. 469-471, 1901. 



