920 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [42] 



about the same in two so nearly related families as Unchytrceidce and 

 TuhificidcB, the much more so because if we consider the same devel- 

 opment of this organ to take place in both families it would at once 

 explain the invagination of the oviduct in a most easy and, as it would 

 seem, most natural way. The oviduct, once formed by an inturning of 

 the body wall, would easily have been invaginated by the efferent duct, 

 originated on and projecting from the septal tissues of a neighboring 

 segment. Its exterior end, instead of connecting with the body wall, 

 would then merely remain suspended in the oviduct, or perhaps even 

 later be connected with this organ through accessory muscles. 



However plausible such a theory may seem, and however easily it 

 may explain one of the most complicated anatomical facts, my own 

 observations have necessitated me to reject the same; and I will in the 

 following endeavor to demonstrate the course of development of the 

 copulative organs in Tuhificidce. The species I have studied for this 

 l)urpose were Limondrilus alpestris, L. coralUnus, and Telmatodrilus. 

 The first named of this species was especially favorable for observation. 



The first sign of the efferent duct I find to be a small glandular, or. 

 at least cellular, agglomeration, situated on the body wall of the tenth 

 setigerous segment, one on each side of the ventral nerve, and exactly 

 on the place where in a future stage of development the genital porus 

 will be found. (Figs. 18 a and b.) 



This cellular agglomeration is convex, and in shape somewhat resem- 

 bling a raspberry. When viewed from above it will be found to con- 

 sist of two different layers, one exterior or cortical (Fig. 18 a: cr. 1.), and 

 one interior (in I). In the same specimens I looked in vain for any 

 aeglomeration of cells on the septal tissues which could be considered 

 as the first beginning of the efferent duct. 



In a further advanced stage this primary agglomeration is found to 

 have increased in size towards the perigastric cavity, and assumed the 

 forui of an oblong body, tapering slightly towards its free interior end. 

 Both the interior and exterior layers are now more differentiated, the 

 former one being more or less transversely striated and the latter seem- 

 ingly composed of longitudinal tissues (Fig. 18 c). The cortical layer is 

 also seen to be covered by a minute coating of pellucid cells, and the 

 interior cellular mass is pierced by a narrow lumen or canal, the future 

 seminal duct. 



When this large glandular body has reached a certain size or a cer- 

 tain stage of development, it opens at the top and the interior matrix 

 grows out and forms a long and narrow tube, the future efferent duct 

 |)roper. 



In more advanced specimens this tube is found to be connected with 

 the septal tissues, and its interior end furnished with a large circular 

 body, considerably flattened, and composed of large round cells. This 

 round body is situated on the other side of the dissepiment, and evi- 

 dently the first beginning of the efferent funnel (Fig. 18 /). The way 



