THE EMBRYONIC DEVELOPMENT OF THE SALPIDAE. 447 



inent, is nevertheless very remarkable, this organ having been con- 

 sidered as the typical feature of the Tunicate plan of organisation. 

 To explain this fact, we might be tempted to assume that there are, 

 among the Tunicates, forms only slightly removed from some primi- 

 tive ancestor which did not possess a chorda, and in which, conse- 

 quently, the development of this organ is not sufficiently established 

 as a constant feature.* 



SALENSKY holds that a rudiment of the larval tail is found in the 

 .Salp embryo in the elaeoblast, that problematical provisional or^an 

 (p. 432) to which TODAKO (No. 107) no doubt erroneously attributes 

 .such a great significance in connection with the development of the 

 prolif crating stolon. The significance of this structure as a vestige 

 of the larval tail seems to be rendered somewhat probable by a 

 comparison with the tailed embryos of Doliolum. [During the 

 development of this organ, however, there is no suggestion of an 

 entodermal origin, if anything an ectodermal one is suggested. 

 Functionally, the elaeoblast appears to be nutritive.] 



Turning to the rudiments of organs, we must first trace the develop- 

 ment of the respiratory cavity. In this we can always distinguish 

 two separate cavities divided from one another by the gill which runs 

 between them obliquely (Fig. 224, k). The anterior and ventral 

 cavity known as the pharyngeal cavity (ph) is considered as the 

 equivalent of the respiratory cavity of the Ascidian, while the pos- 

 terior cavity which lies dorsal ly to the gill is regarded as the atrial 

 cavity (cl). The two large apertures through which these two 

 cavities communicate on either side of the gill are regarded as un- 

 usually dilated gill-clefts. In this view, which is favoured by the 

 condition of the gill in Doliolum, this one pair of gill-clefts occurring 

 in the ticdpidw has been thought to arise by the fusion together 

 of several smaller clefts. This view is opposed by VAN BENEDEN and 

 JULIN (No. 10), according to whom only one pair of true gill-clefts 

 develops in the Ascidians also, the many perforations of the wall of 

 the gill which develop later being secondary structures (branchial 

 .stigmata, p. 367). The Salpidae in this case would, like the Larvacea. 

 exhibit a very primitive character in the presence of a single pair 

 of gill-clefts. TODAKO (No. 113), who adopts this view on the whole, 

 has extended it by explaining certain ciliated invaginations which 

 are found arranged in rows at the sides of the gill in some Salps 



* We must, however, bear in mind that the body of theThaliacean represents 

 principally the pre-chordal region of the Ascidian larva. 



