Boyal Microscopieal Societij. 269 



a one is represented in Fig. 3. It consists of a pedicle, &, attached 

 to a dome-like depression on the under side of the tarsal joint, 

 a, c, a, and a disk, d, e, f. The npper surface of the dome-Hke 

 depression forms the elevation already alluded to as a nipple-shaped 

 projection into the interior of the tarsus. The secreting sac is 

 closely united to its margin and is represented at li, h ; its apex is 

 perforated at c. 



The pedicle consists of an elastic external membrane, h, thick- 

 ened into a chitinous ring around its junction with the disk, and 

 lined with epithelium. It is hollow, and strengthened by a hollow 

 columella, c, which is open along one side of its whole length. The 

 disk is a flattened membranous sac, continuous with the outer 

 membrane of the pedicle above, and in this it difiers only in size 

 from the hairs already described from the fly. The disk is strength- 

 ened, however, by numerous radiating fibres between its layers, 

 which arise from the columella. This consists of a hollow bundle 

 of chitinous fibres, which are attached above around the opening in 

 the dome, c, and below by a sjoeckled cone of chitine, g, to the 

 centre of the disk. The fibres pass over or through this cone in a 

 radiating direction, and become applied closely to the lower mem- 

 brane of the disk. Here they branch dichotomously, and ultimately 

 unite, at about one-thuxl the diameter of the disk from its centre, 

 with both membranes. They then run straight to the margin of 

 the disk, leaving channels between them, through which the viscid 

 secretion from the tarsal sac passes. This secretion fills the cavity 

 of the hair, which it enters through the columella. The extremi- 

 ties of the radiating fibres on the margin of the disk, invested by 

 a prolongation of its wall, form a dense fringe of brush-like tufts 

 around its circumference. An ovate space on the surface of the 

 disk (Fig. 4, a) may be observed destitute or nearly destitute of 

 radiating fibres : it corresponds to the open side of the columella. 



I have been unable to discover how the secretion is prevented 

 from flowing from the disks when they are not in use, but I suspect 

 this is effected in the following manner. When the disks are at , 

 rest they are cup-shaped ; when they are applied to any flat surface 

 they become flattened out. The cup-shaped form of the disk is 

 maintained by the elasticity of the outer wall of the pedicle. This 

 keeps the columella stretched, and I suspect keej^s the slit in its side 

 closed. As soon as the disk is flattened the tension is taken off" 

 the columella, and I suspect its slit gapes just as a bundle of strings 

 kept together by tension upon their ends separate when held loosely. 



The smaller of the large disks is exactly like the larger, excejDt 

 in size. The small disks differ, however, from the large ones in 

 having no columella and no fringe. The pedicle is difiierently 

 shaped, being pointed at its junction with the disk. The radii of 

 the disk are very delicate, and terminate in the centre of the 



