GILL DEVELOPMENT IN MYTILUS. 63 



Ostrea has been briefly considered by Ryder ('84) and Stafford 



Co5). 



Mytilus has been equally briefly described by Wilson ('87). 



Dreissensia is known through the work of Korschelt ('91), 

 Weltner ('91), and Meisenheimer ('01). 



And, finally, the gill development of the peculiar and primi- 

 tive Nuculidae, Nucula and Yoldia, has been very carefully fol- 

 lowed by Drew ('99, '01). 



The accounts are for the most part brief, in some cases only 

 incidental ; and there are various points of divergence in detail. 

 One such divergence appears at first to be fundamental. While 

 Lacaze-Duthiers describes the first anlage of the gills as a series 

 of isolated papillae, the majority of the authors cited trace the 

 ctenidium back to a longitudinal fold or ridge, which is second- 

 arily constricted transversely and divided into the papillae. Thus 

 emphatically, among later writers, Ziegler for Cyclas, Schierholz 

 for Unio, and Sigerfoos for Teredo. Drew describes a primary 

 ridge and secondary papillae in the Nuculidae, but considers the 

 change to be " due to unequal growth more than to constriction " ; 

 his figures, however, give evidence of considerable constriction. 



On the other hand, Wilson confirms the early view of Lacaze- 

 Duthiers for Mytilus ; and my own observations are entirely in 

 accord with this view. Concerning Dreissensia Korschelt ('91, 

 p. 144) expresses himself thus cautiously : " Ob sie (the gills) 

 in Form einer Falte angelegt werden, die sich schon sehr bald 

 einkerbt und so jene vermeintlichen Papillen entstehen lasst, oder 

 ob sie als wirklichen Papillen hervorsprossen, ist schwer zu 

 entscheiden." 



In the adult of all forms a continuous and undivided gill axis 

 may be recognized, comparable with this embryonic fold or ridge. 

 Upon this are carried the filaments, comparable with the embry- 

 onic papillae. The whole divergence, then, reduces itself to a 

 question of the relative time of development, and is one of minor 

 importance. Much more significant is the general uniformity of 

 development, and the recognition, by all the authors cited and in 

 all the genera studied, of a stage in which each ctenidium is rep- 

 resented by a series of simple papillae belonging to the inner gill ; 

 and of a later stage in which the outer gill is formed of a par- 

 allel series of corresponding papillae. 



