1919] Metcalf — Respiratory Apparatus of Dipterous LarvoB 55 



factorily used, and in this case the easiest measurement to make 

 appears to be the maximum, strictly transverse diameter.^ 



(6) // stalked, the length of the posterior respiratory organ or tube: 

 the elevation of the stigmal plates above the general surface of the 

 segment which bears them. 



(7) The height of the stigmal plate or respiratory organ: the maxi- 

 mum dorso- ventral (sometimes apparently cephalo-caudal) meas- 

 urement. 



(8) The presence or absence of the circular plate or button. (Banks.) 



(9) The position of the button, whether on the circumference of the 

 stigmal plate ("ring" of MacGregor) or within it. (MacGregor.) 



(10) The diameter of the circular plate or button: a strictly trans- 

 verse measurement through its center. 



(11) The distance bettveen the mesal margins of the two circular 

 plates.- 



(12) The length of the slit-like spiracles. 



(13) The width of the spiracles. 



(14) The shape of the spiracles, which are for convenience of 

 further description named dorsal, median and ventral. It is 

 necessary to note that the several spiracles on the same stigmal 

 plate may vary' greatly in shape and it is often necessary to refer 

 specifically to them.^ See figure 1. 



(15) The ornamentation of the stigmal plate with spines, tubercles, 



1 Professor Herms {loc. cit.) has used " the diameter of the stigmal plate, the space occupied 

 by one stigmal plate on a line drawn through the center of both; (2) length, when slits are absent, 

 the space occupied by a plate on a line drawn dorso-ventraUy through the center of the plate; 

 or when slits are present the space occupied by a plate along a line drawn from the lower edge of 

 button (or space if button is absent) through the longest slit (middle slit) to the margin of the 

 plate; (3) width, along a line drawn at the middle of the plate at right angles to the length 

 line." These are so defined as to be a little difficult to measure, — requiring an estimation of 

 such features as "the center of the plate," and a line at right angles to another. They are to 

 this extent apt to lead to error, and will, I oelieve, be adequately supplanted by the more readily 

 determinable features suggested above in paragraphs 5 and 7. 



' MacGregor suggests measuring from center to center of buttons, which complicates the 

 matter by necessitating an estimation of the central point of each plate. Point 11, in connec- 

 tion with point 10 will give the same result more easily. 



' MacGregor {loc. cit.) attempts to divide the forms of posterior stigmata into two main 

 types: the schizotreme-type especially characterized by possessing three pairs of slit-like spira- 

 cles; and the plychotreme-type, possessing instead " a convoluted chain" (e. g., Muaca domestica, 

 Stomoxys calcitrans, Hamatohia serrata, etc.). But he, himself, points out that there are usually 

 two ("three," loc. cit., p. 181) breaks in the chain. I think there is no sound basis for such 

 distinction. Both of these forms have three pairs of spiracles; the convoluted spiracle and the 

 slit-like one are fundamentally homologous, differing only in shape; and all possible intermediate 

 grades of convolution and complexity are to be found. In cases of Tery great complexity, it is 

 true, the spiracles may run together so that it is difficult to determine the limits of each, but I 

 believe such forms have the same origin. 



