BDELLOSTOMA DOMBEVI, LAC. I 39 



wards. These simple primitive slits increase in number until 

 about 14 of them are laid down, when a stage is reached during 

 which no more slits are formed in the antero-posterior series. 

 This stage has been designated the critical stage by Willey. 

 When the activities of growth again manifest themselves in 

 these organs it is an entirely new fashion and results in the 

 splitting of each primitive gill-slit into two by the downward 

 growth of a uicdian gill-bar which thus separates an anterior 

 from a posterior portion of the original aperture. After this 

 process has begun, increase in the number of gills from before 

 backwards again takes place ; but is now evidently related to 

 the acquired habits of the animal, as I have already pointed 

 out in another place. ^ In the matter of gills, we see that 

 Bdellostoma approaches very close to Branchiostoma, and in 

 this sense the Myxinoid retains very primitive relations of its 

 gill-apparatus. To return to the question of the variability of 

 the gills in Bdellostoma, what can be its significance } Have 

 we to do with an increase or decrease in the number of gills of 

 this animal, seeing that by far the largest number of individuals 

 now existing have either 10 or 12 gills on both sides of the 

 body .^ I think there can be no doubt as to the answer. We 

 are here dealing with a reduction in the number of gills and in 

 no case with an increase. This conclusion is supported by the 

 fact that the more highly differentiated Marsipobranchs have a 

 smaller number of gills, and this is true not only of Myxine, 

 which belongs to the same group as Bdellostoma, but in a 

 greater degree of the Petromyzonts. In view of these facts it 

 is extremely probable that the ancestors of Bdellostoma were 

 provided with about 13 to 14 gills, and this number may be 

 taken as representing the numerical relations of the branchial 

 apparatus of the primitive forms of the vertebrate stock 

 generally. The relation of the base of the tongue-muscle to 

 the gills is of interest, and here again we find great variability. 

 Muller found it to lie entirely in front of the gills in the 6 and 

 7-gilled forms from the Cape of Good Hope, and this condition 

 obtains in Myxine so far as known. In Bdellostoma with 10 

 or 1 1 gills the base of this muscle may lie between the 6th and 



^ Concerning Vertebrate Cephalogenesis. — Jour)ial of Morpholof^y, IV, 1890. 



