THE STRUCTURE AND AFFINITIES OF HERDMANIA CLAVIFORMIS. 255 



V. SOME POINTS IN THE EMBRYOLOGY OF HERDMANIA. 



The developmental history of this new form lies beyond the purpose of this paper. 

 It may not be amiss, however, to summarize a few of the more important results 

 thus far reached by my studies under this head. 



1. After cleavage begins, the ova are surrounded by a very large number of test- 

 cells (PI. XVIII, Fig. 8, cl. 1st.). These take no direct part in the production of the 

 embryo. They are perfectly distinct from the tissue of the embryo throughout their 

 whole existence; and their gradual disintegration and resorption can be easily fol- 

 lowed. 



2. The peribranchial sacs arise as two very long, narrow, ectodermal invagina- 

 tions on the dorsal side of the embryo. The endoderm does not contribute at all to the 

 formation of these sacs, or only to the slightest extent. 



3. A single pair of very large, nearly round, protostigmata are the only com- 

 munication between the branchial and pouch-like peribranchial organs for some time 

 before any of the definitive stigmata are formed; so that the Appendicularia stage 

 in the ontogeny is unusually clear. The definitive stigmata all arise by perforations 

 wholly independent of the protostigmata. 



4. The origin of the ganglion and neural gland from the common ectodermal 

 anlage can be followed with great ease and distinctness. 



5. The entire embryonal and larval development are passed within the uterus. 

 The larval period is greatly abbreviated, and there probably is no free-living larval 

 stage. 



VI. AFFINITIES. 



Having now obtained a somewhat detailed acquaintance with the structure of 

 the new ascidian we are in a position to devote with profit some attention to ihe 

 question of its affinities. To facilitate this I arrange its leading characteristics in 

 tabular form on the basis of their similarity to those of the nearest allies. The types 

 of three families must enter into the comparison: Amaroucium (Polyclinida?), 

 Clavelina (Clave linidse), and Distoma (Distomidse).* 



* Polyclinopsis of Gottschaldt appears to have some interesting resemblances to Herdmania. Concerning 

 the embryos Gottschaldt ('94, p. 355) says " dass bei jiingeren Personen sich ein Tubus dorsal direkt unter dem 

 Ektoderm durch das Postabdomen hinzieht; derselbe ist mit Eiern angefiillt, die vorzugsweise hintereinander 

 liegen und jugendliche Stadien zeigen. Bei fortgeschritteneren Personen finden sich dichte Eihaufen, rings das 

 Entodermrohr einschliessend und dasselbe teilweise einengend, am oberesten Teile des Postabdomens in alien 

 Stadien der Entwickelung (s. Taf. XXV, Fig. 3); bei anderen liegen dieselben Eimassen mitten in Kiemondarm. 

 Hieraus kann man wohl, ohne einen Fehlschluss zu thun, folgern, dass die im Postabdomen hereangereiften Eier 

 sich am oberen Ende desselben sammeln, in den Kiemendarm eindringen und diesen als Brutraum benutzen." The 

 nature of the "Tubus" here mentioned is a matter of much interest. One might suppose it to be the oviduct did 

 not the author expressly state on the preceding page that this duct seems to be wanting. 



