AND TORNARIA. 433 



terior extremity, and disappear entirely at a considerable distance behind the collar 







on the dorsal side. The median dorsal part of the body beyond the gills is somewhat 

 flattened, becoming more so for the greater part of the distance occupied by the liver, 

 the position of which is somewhat different from that which it occupies in B. clavi- 

 gerus, where it extends only a short distance below the gills towards the collar ; the 

 lateral folds do not take a great development, and do not unite dorsally behind the 



| 



collar, as is the case in B. minutus. The whole surface of body, as has been mentioned 

 by previous writers on Balanoglossus, is thickly covered by minute cilia. 



The gills, as I have shown from their mode of formation, consist originally of folds 

 of the oesophagus, forming subsequently elliptical, funnel-shaped diverticula from it ; 

 from the dorsal sides of these, new folds are formed, dividing the funnels into two; 

 and so additional folds are formed, increasing greatly the complexity of the gills, but 

 never, in our species, forming the remarkable system of network described by Kow- 

 alevsky; nor have I been able to make out any special order in the mode of 

 formation of the folds of the gills. Their mode of opening externally is quite 

 different from that described by Kowalevsky. Near the dorsal median line we find 

 a series of inverted pouches which the slightest compression will throw out like an 

 inverted finger of a glove, forming a flat cylinder opening into a narrow slit next to 

 the dorsal vessel, through which the gills communicate externally. The inner walls 

 of this cylinder are strongly ciliated. The supports of the gill-folds are quite simple ; 

 the folds of the gills are supported by three prongs starting from a common curved 

 base and attached to the more or less granular chord extending between them along 

 the dorsal line ; there is nothing to be seen of the complicated skeleton support of 

 the gills figured by Kowalevsky for B. minutus. It is immediately on the edge of 

 the folds that the most powerful vibratile cilia are found ; owing to the increased 

 lengthening of the central and lateral folds of the gills, they occupy a greater part of 

 the gill-opening, and, becoming laterally crowded, appear like numerous folds placed 

 side by side, while in reality we see only the edges of the folds, and of their skeleton 

 supports in profile. 



The genital organs occupy the same position as described by Kowalevsky for the 

 Neapolitan species, between the liver and the anterior part of the body, forming 

 singular bags on either side of the median line. Only a few eggs were found, and 

 all attempts to raise them by artificial fecundation failed completely. Nor did I suc- 

 ceed, while digging over a large extent of ground occupied by Balanoglossus, in 

 finding any trace of strings of eggs, as Kowalevsky suggested they might be found. 

 The posterior part of the body is quite cylindrical, the alimentary canal having but 



