274 BULLETIN OF THE 



The most peculiar structure in the anatomy of the knob is the prolongation 

 of the margin of the involucrum on its lower side into a conical appendage (a), 

 which extends out from the knob slightly beyond the vesicle. This conical 

 body appears to be solid, and is crossed by lines or annulations, as shown in 

 ft"S. 9, 10. It has a yellow color, and may be called the apex. 



If now we compare the first kind of knob in A. formosa with that of A. 

 rosacea, Esch., we find this important difference between them. In rosacea, 

 according to Kolliker, there is a " stalked elongated capsule " banging to the 

 knob at its base, near the origin of the involucrum. That capsule is undoubt- 

 edly homologous to the apex (a) of the knob in formosa. Its closest homology, 

 however, is with the " second kind " of knob in A. formosa, as we shall show 

 presently. 



Sars * has already compared the knob of A. rosacea with that of Agalma. 

 The resemblance of the first kind of pendant in A. formosa to that of Agalma 

 is even more striking, since it possesses an involucrum of a form which has not 

 been observed in A. rosacea. 



A second kind of tentacular knob (figs. 7, 8) is sparingly scattered along the 

 tentacle of A. formosa. Besides being less numerous, it is also much larger 

 than the former, with which it is homologous. It may eventually be found to 

 have been developed from the preceding. This knob has one of the most 

 peculiar forms which this organ, highly variable among Siphonophores, 

 assumes. Its general shape is shown in figs. 7, 8. 



The knob, when seen from the side (fig. 8), has a pear shape, and hangs 

 from a short flexible peduncle. The great mass of the pendant is made up of 

 a very large involucrum or structure of the same homology. The walls of this 

 body are very thick, and seem to be composed of large cells. At one end it is 

 continued into dendritic branches (a a) of a yellow color. The body of the 

 involucrum is transparent, and through its walls there can be seen a crimson 

 and orange-colored sacculus (//)• The sacculus is not coiled, as is generally the 

 case with other Physophores, but has a simple curved shape, and in its walls 

 can be traced the rows of large thread-cells, which are a common feature of all 

 these organs. 



Fastened to the upper side of the involucrum, near its articulation with the 

 peduncle, there spring three bodies very similar to those found on tl. 

 kind of knob. One of these structures is the ovoid vesicle (<), and on tech 

 side there rises a short lateral filament (b). These organs differ in no r- 

 from the same in the former kind of pendants. 



The most exceptional feature in the second kind of knob in Athorybia i 

 in the structure of the distal extremity of the involucrum or the apex (<()• 

 The apex (fig. 8, a a) is bifid ai its base, and after a basal bifurcation each part 

 subdivides into many .-mailer unbranched divisions. Kadi division lias the 

 general appearance of the apex of the oral kind of knob. It is capable of grea^ 

 extension and retraction, and when drawn hack baa a corrugated surface, like 

 similar branches in the " Handlbrmi. • - Angelorgan," described by Qegenbaux 



* Op. cit. 



