338 ARTHUR WILLEY. 



homology of the first pair of gill-slits in all three cases, 

 whether permanently, temporarily, or never functional, is no 

 less clearly demonstrated. Dohrn's argument, therefore, 

 against the primitive character of the Ascidians, based on the 

 supposition that the pericoronal groove was the remnant of a 

 pair of gill-slits, falls entirely to the ground. 



It will probably come to be recognised that the first pair of 

 gill-slits of the Ascidians, Amphioxus, and Ammocoetes is in 

 every sense of the word, morphologically and actually, the first 

 pair of gill-clefts, there being absolutely no evidence of the 

 mouth and endostyle of the Ascidians and Amphioxus having 

 each proceeded from a pair of gill-clefts, but a great deal of 

 evidence to the contrary, as we have seen and shall see. 



It is hardly necessary to add that the prodigious degenera- 

 tion postulated by Dohrn for the Ascidians and Amphioxus 

 cannot be accepted in the light of the above considerations. 



5. The Atrial Cavities of Ascidians. 



After what has been said above as to the primary positions 

 of the several organs relative to the endostyle in the larval 

 Ascidian and Amphioxus, the homology which van Beneden 

 and Julin sought to establish between the primary atrial 

 cavities of the Ascidians, and the two anterior preeoral 

 intestinal diverticula of Amphioxus, will appear almost in- 

 conceivable (see PI. XXX, figs. 1 — 5, for position of atrial 

 apertures). 



The idea of a prechordal vesicle, which was the starting- 

 point of the above-named authors in their comparison of the 

 Ascidians with Amphioxus, and led them to a remarkable 

 series of positive and negative homologies, and which in 

 ignoring the presence of a very large and prominent praeoral 

 lobe, which we have shown to occur in the Ascidians, seemed 

 plausible enough, is undoubtedly in its main features a mis- 

 conception. 



From observations made on Ciona intestinalis and 

 Clavelina lepadiformis I can confirm Kowalevsky (21) and 

 Seeliger (31) iu saying that tlie atrial cavities of Ascidians 



