248 THE STRUCTURE AND AFFINITIES OF HERDMANIA CLAVIFORMIS. 



The origin of the tubes in Herdmania, then, agrees perfectly with what is known 

 of them in Clavelina and the merosomatous * ascidians. The presence of two epi- 

 cardiac tubes in full development is not the least among the many particularly 

 interesting facts in the structure of this unique ascidian. 



There is no known species in the whole range of bud-producing tunicates quite 

 like the present one. The tubes, where they occurred, were indeed primitively double 

 without much doubt in all ascidians, but in all without exception, so far as we now 

 know, one of the tubes has been more or less completely merged into the other, so 

 that for most of the length only a single tube is present. In all cases, probably, the 

 two are distinct at their origin; but they everywhere fuse to produce the epicardiac 

 sac (Van Beneden et Julin, '86). 



In Fragaroides (Maurice, '88) and probably in the other Polyclinidse the double 

 character of the tube appears again at its posterior extremity where it forks, one 

 limb being dorsal and the other ventral to the pericardium. Herdmania has in this 

 respect, then, more in common with the Polyclinidse than with any other group. 

 In both, the heart and pericardium are situated far back in the postabdomen, and 

 in both the pericardium is placed between the ends of two epicardiac tubes. t But 

 in Fragaroides and all other Polyclinidse the tubes become fused together a short 

 way above the pericardium, and continue in this condition almost up to the 

 pharynx. 



Could we interpret the posterior extensions of the peribranchial sacs of Botryllus 

 as epicardiac tubes, as Pizon ('93) and Garstang ('95) have done, we should have 

 a sort of agreement between these two genera. But both Hjort ('96) and Ritter 

 ('96) have shown that these structures in Botryllus cannot be homologized with the 

 epicardia. The BotryUidaB and Polystyelida? in reality have no epicardiac tubes. 



Furthermore, the raphe of the heart in Herdmania, as in the Polyclinida3, is not 

 in such relation with the wall of the epicardium as to be closed by it as is the case 

 in Clavelina. J 



It is possible to institute a suggestive comparison between the stolon of Herd- 

 mania and that of Doliolum, though no great confidence can be placed on the com- 

 parison until we have more exact knowledge about the origin and prospective value 

 of the various elements that enter into the structure of the stolon of Doliolum and 

 until we also know how the bud originates in Herdmania. I consequently merely 

 indicate the homologies that seem most suggestive. The comparison which I make 



* This term, proposed by Sluiter ('95) and adopted with some modification by Herdman ('99), is a very useful 

 one. 



t Compare Figures 9, 10 (PI. XVII) and Figure 11 (PL XIX) with Maurice's ('88) Figure 70 (PI. XIX). 

 t Compare Figure 10 (PI. XVIII) and Figure 11 (PL XIX) with Figure 66 (PL X) by Van Beneden et Julin ('86). 



