128 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



The circuUir :iikI outer longitudinal rings of muscles have been greatly reduced in this cirrus, 

 the longitudinal having almost completely disappeared. Immediately underneath the glands 

 is a thin layei- of the circular muscles. With the development of the glands the dermis tissue 

 has bei'ome greatly increased, so that tiow the glands are surrounded by tibrous tissue, traversed 

 here and there by muscle tibres. Each gland has a thin tunic of muscle fibres. A few vascular 

 lacuna- are found in the dermis. 



The glands have the shape of thick, round-bodied flasks with very short and narrow necks. 

 (Fig. 55.) Owing to their mutual pressure the glands are usually polygonal. The}' are 0.8 to 1 

 millimeter in length and 0.4 to 0.<3 millimeter in diameter. The exterior of the cirrus is covered 

 by an epithelium of very slender columnar cells, none of which are secretory. At the mouths of 

 the glands a transition occurs to shorter or even cubical cells which line the duct and upper part 

 of the body of the gland. The sides and the base of the gland are lined with exceedingly long, 

 slender, secretory cells, whose oval nuclei lie in the very bases of the cells. There is considerable 

 variation in the length of the cells of the upper side of the gland. Sometimes they are scarcely 

 longer than those of the duct; sometimes, again, as long as any in the gland. Mucous cells are 

 sometimes scattered among the cells of the upper sides and the duct of the gland, the other 

 cells of these regions not appearing to be secretory; again, all are secretory. The portion of 

 the secretorj' cell above the nucleus is closely packed with fine granules. The lai'ge lumens of 

 the glands were in most cases tilled with a finely granular secretion. Glands near the tip of the 

 cirrus may be more or less distinctly divided into two portions, which are partially separated by a 

 constriction — a basal thick-walled secretory portion and an outer thin-walled portion, which may 

 be a collecting or storing chamber, from which the duct leads to the exterior. 



The ridges upon the surface of the cirrus are almost entirely occupied })y large \ascular 

 lacuna?. A network of connective tissue extends between the lacunte of the ventral ridges, tiut 

 there is almost no connective tissue in the ridges of the dorsal side. 



THIRD CIRRUS OK SI'ADI.X. ( FIGS. 17 AND 18.) 



The third cirrus is the most highly modified and remarkable of the group. It is longer and 

 larger than the second, being intermediate in size between this and the first. As has ahcady 

 been noted, its l)ase lies posterior to that of the second. It is oval in ci'oss section at the liase 

 instead of round, as is the case with all the oth(>r cirri of the spadix. It is narrowed at the base, 

 however, like the second cirrus. For al)out half its length it is oval and perfectly smooth. At 

 its middle it begins to be still more flattened dorso-ventrally, until near its tip it is almost per- 

 fectly flat upon both sides. The dorsal side is apt to be somewhat concave and the ventral side 

 convex. A longitudinal ridge may be present along the median line of the dorsal side near the 

 tip, caused by the tissues passing over the nerve lying directly underneath. 



At the same time tliat th(! cirrus becomes flattened it loses its smooth surface. The dorsal 

 surface becomes marked l)y transverse ridges, which are at first verv faint, but become distinct 

 as the tip of the cirrus is approached. At the very tip, however, the ridges become crowded 

 and less distinct. They also disappear at the edges of the cirrus, leaving a smooth, unmarked 

 margin. Close examination with the nsiked eye discovers many smaller ridges extending longi- 

 tudinally between the transverse. 



It is upon the opposite (the ventral) side of the cirrus that we find a most peculiar structure. 

 (Fig. 18.) Arranged across the cirrus in rows which seem to correspond closelj' to the ridges 

 of the opposite side are great numbers of fine pits extending into the tissue of the cirrus. The 

 largest of these may be a little more than a half a millimeter in diameter. Examination with a 

 hand lens shows the openings to lie nearly scjuare, the sides being parallel to the main axis of the 

 cirrus, and to l)e set as closely to each other in both directions as is possible. The openings are 

 often so regularly spaced that they form longitudinal as well as transverse rows. Only a thin 

 wall remains l)etween the pits. At the jjosterior end of the pitted surface the pits ))ecome small, 

 imperfectly formed, and finally disappear. Similarly the pits are not well developed at the tip 

 of the cirrus. When the fingers are passed over the pitted surface it feels like shagreen, and 



