518 STUDIES IN LIFE-HISTORIES OE AUSTRALIAN DIPTERA BRACHYCERA, i., 



twelfth segint'iits, and the jjositidu of the posterior spiracles would eorrespoiid with 

 that in which they are usually fouud in dipterous larvae ha^^ing twelve segments. 



Pupal metamorphosis. 



In this paper I do not propose to do more than indicate in a general way 

 the external features of the development of the pupa. 



Jusbaschjanz (1910) has studied, in great histological detail, the metamor- 

 phosis of various internal organs in the larva, but says very little about the 

 pupa, and, so far as 1 am aware, no description exists of the stages in the 

 gradual change of form during its growth from the larva. 



As the pupa is hidden inside the old larval skin during the whole course of 

 its development, it is necessary to remove the skin in order to study it. At a 

 very early stage in its metamorphosis, the pupa comes to lie free inside the 

 skin, surrounded by a watery fluid, and retaining its connection with the skin 

 only by means of the stigmata on each segment Its removal, therefore, is an 

 easy matter, and is rendered still easier by the existence of lines of weakness in 

 the skin at the points where it is subsequently split open by the emerging fly. 

 Old, empty larval skins show the lines of cleavage very well. By a clean, 

 circular cut round the upper part of the second thoracic segment, the portion of 

 the case anterior to this is separated ofi in the form of a sort of "head-cap.'' 

 In the mid-dorsal line, the circular cut dips to form a sUght angle, and from 

 this point a straight split extends down to the ujiper part of the first abdonunal 

 segment, wbei'e it meets a second transverse split extending nearly across the 

 full width of the dorsal wall. After the emergence of the fly, the edges of these 

 splits fit closely and evenly together, owing to the rigidity of the walls, so that 

 the skin has the appearance of an entire ease, with a circular aperture at one 

 end (Text-fig. 4) . ' 



These natural lines of cleavage are found to be present from the earliest 

 stages of pupation. The "head-cap" is easily removed by light pressure with the 

 point of a needle^ and a similar pressure opens up segments two to four, mid- 

 dorsally, in a longitudinal direction, then circularly round the fourth segment 

 (Text-fig. 16), so that this portion of the skin, forming a "thoracic band'" can 

 be stripped off in a single piece. By carefully cutting along the mid-dorsal line 

 of the abdominal larval skin, the pupa can be removed, still enclosed in a deli- 

 cate transparent Dupal membrane. 



In the earliest stage of a female pupa removed in this way (PL xxvii., fig. 3) 

 pupal head and thorax have become differentiated, but the abdominal segments 

 still retain the larval shajie. and differ from those of the larva only in the char- 

 acter of tlie integument, and in the presence of stout, projecting spiracles with 

 brown chitinised tips. With this exception, the entire pupa is very soft, and 

 pure white. No trace remains of the hexagonal armoured plates of the larval 

 skin, or of the bristles, or other integumental structures. The enclosing mem- 

 brane fits closely over the body, and, except where it is raised into "blisters" 

 over developing appendages, it is not apparent. 



The position of the mouth parts is indicated by blister-like swellings of this 

 character. Already well defined at this early stage, they form, at later stages, 

 much more conspicuous features of the pupal head, than do the moutli parts on 

 the adult. 



The iliree tiioracic segments arc still distinct, and the appendages of each 

 are folded closely against its ventral surface, and do not extend beyond the seg- 



