28 ME, M. LAUKIE ON THE 



structure and appearance are precisely similar to what has 

 already been described. This asymmetrical arrangement of what 

 one must regard as a morphologically symmetrical structure 

 is interesting on account of its rarity. The Arthropoda are 

 essentially bilaterally symmetrical animals, and yet here we have 

 a bulky organ disposed in a completely unsymmetrical way, and 

 that without appreciably affecting any of the other organs in the 

 same region of the body. The extreme ventral position of these 

 sacs is also worth noticing, as it is but seldom that any structure 

 of importance comes to lie between the nerve-cord and the 

 ventral surface. 



Twisting about on both sides of the central or right stink-sac, 

 but more especially on the right side, where there is more room, 

 is a convoluted mass of fine tubules. The convolutions are so 

 complicated and the tubules so fragile, that I have not been able 

 to ascertain how many tubules are present or whether they 

 branch or anastomose. This last is probably not the case, as I 

 could scarcely have failed to get some trace of branching if it were 

 present. I have traced two of these tubules apparently opening 

 into the distal, i. e. anterior, end of the left stink-sac, and have 

 little doubt that others open similarly into the right one. I take 

 these tubules to be the purely secretive part of the stink-gland, 

 and imagine that they discharge their secretion into the sacs, 

 from which it is ejected in considerable quantities when necessary. 



The coxal gland (PI. III. figs. 2 and 4, cox?) lies in the thorax on 

 either side of the entosternite, the processes of which pass dorsal 

 to it. It is an elongated body with a wavy outline, and the con- 

 volutions of the tube of which it is composed may be seen on 

 the surface. At the front end it gives off a duct which runs 

 alongside the foremost process of the entosternite, and then 

 curving outwards passes into the base of the third appendage, 

 I have been quite unable to find any aperture on the external 

 surface in this region, but there is a considerable membranous 

 area in which such an aperture might easily be overlooked. At 

 the same time it is quite possible that the duct may be closed in 

 the adult, or only open at special seasons as in Mygale. 



The lung-booJcs are situated, as has been already stated, towards 

 the sides, beneath the first and second abdominal sclerites. The 

 lamellae lie for the most part horizontally, though curving up a 

 little towards the outside. Each lamella has a comparatively 

 short posterior edge where it abuts on the air-space (fig. 8). 



