ON THE DEVELOPMENT OF LINGULA ANATINA. 103 



was clearly observed by Kowalevsky in oistella. The authors 

 who insist from anatomical grounds that there is a close relation- 

 ship between the Brachiopoda, Polyzoa and Phoronis, or who 

 acknowledge that the Brachiopoda form part of the " Trocho- 

 zoaires " from the similarity in structure of their larvse, have 

 always been perplexed by the enterocœlic type of the mesoblas, 

 formation as stated by Kowalevsky. From this single instance 

 of Cistella the Hertwigs f8i) maintain that the Brachiopods 

 have the nearest affinity to the Chœtognaths and their view has 

 been supported with several facts by van Bemmelen ('83). R. 

 Hertwig is still of this opinion {Cf. '00 p. 298). Boule ('94) 

 indeed goes so far as to doubt the observations of KowALEvsKYt 

 citing his similar observations in Phoronis to show how apt we 

 are to fall into mis-interpretation when we study " ses trans- 

 formations si passagères " only from surface views. But it is, I 

 think, better not to doubt the facts witnessed by Kowalevsky 

 until we have more exact observations of this point in the same 

 animal. At any rate, according to my opinion there is as little 

 difference between the schizocœl and enterocœl types of the 

 formation of the body cavity (deuterocœl — Ziegler), as between 

 gastrulation and polar ingrowth (Ein Wucherung). It is there- 

 fore nothing to be wondered at if these two modes of body- 

 cavity formation prevail in the nearly allied forms. 



In Lingula the mouth is formed at the place where the 

 blastopore was closed, as in Oistella and Lacazella. While in 

 Lingula the oesophagus and the stomach come into direct contact 

 with both the ventral and dorsal ectoblast ; in oistella, according 

 to Shipley, they touch only the dorsal wall. 



In all Brachipod-forms whose embroys have been studied the 

 larva? are divided by constrictions into three regions or "segments," 



