Structure of Scent Organs in Male Danaine Butterflies. 157 



a more primitive condition, resembling thick scales of 

 irregular section and comparatively great length. They 

 arise from cells with chitinous sockets quite similar in 

 appearance to those giving rise to the other hairs or scales, 

 but above the stalk they are abruptly expanded so that 

 their mean diameter is much greater than that of the 

 sockets from which they arise. The base of the brush is 

 composed of these structures only, and in the retracted 

 condition they form a central cone, lying in the heart of 

 the brush and ending in a point at about two-thirds of 

 the distance from the base of the brush to its extremity. 

 These bodies readily take stains such as haematoxylin and 

 carmine, and are not chitinised to anything like the same 

 extent as other portions of the brush. PI. XVIII, fig. 5, 

 shows a section of one of these structures as seen under a 

 very high power. It is evident that the shape is such as 

 to increase as much as possible the surface area. More- 

 over, I am inclined to think, after careful examination, 

 that numerous pores exist in the surface, so that it is 

 reasonable to suppose either that some secretion is dis- 

 charged therefrom, or that they serve as a store for the 

 secretion extracted from the wing-glands. There is, 

 however, no direct evidence of their function. 



PI. XII, fig. 1, is a photograph of a section of the brush, 

 where it consists solely of these structures. The membrane 

 of the brush-bag is here very thin and contains many small 

 nuclei. PI. XII, fig. 2, shows a section taken a little above 

 the constricted portion of the brush. Here the outer 

 membrane has become thicker, and the large nuclei are 

 those of the ordinary hair-producing cells. Two new 

 structures are now seen to have arisen. 



(1) Forming a ring round the central bodies already 

 described are structures whose walls are evidently thin, 

 and present in section an irregular and considerably 

 convoluted appearance. These bodies enclose and thicken 

 the cone in the heart of the brush, dying out to a point a 

 httle nearer its distal end. A highly magnified section of 

 one of these is shown on PI. XVIII, fig. 6. Inside the thin 

 wall is an exceedingly delicate tissue with distinct dots 

 scattered through its substance, and showing a more or 

 less radiate striation. The walls of these bodies do not 

 show signs of the presence of pores. 



(2) Outside these " first-ring " bodies we see the ordinary 

 brush-hairs — not, at this level, very well rounded in sec- 



