198 University of California Publications. [zoology 



the histological changes in the anterior mesoblastic vesicle 

 leading to the proboscis musculature begins very early. The 

 result of these and other changes which affect the anterior 

 end first and most pronouncedly is that this end probably comes 

 to have a greater instead of a less specific gravity before the 

 swimming life is abandoned, and as a couserpience the orientation 

 is changed, the anterior end becoming directed downward, so 

 that the downward journey is made anterior end foremost, as 

 was the upward journey. 



2. Tornaria liubbardi* n. sp. 



An occasional specimen of a tornaria different from any 

 hitherto described was taken along with T. ritteri at various 

 times during the summer. This new species is particularly 

 distinct from all others known in the large number of gill 

 pockets present before any other indication of metamorphosis 

 appears. At least five pairs of these could be seen on the living 

 animal in the youngest larva taken; and at an early stage of 

 metamorphosis seven pairs were present. 



Another striking characteristic of the species is found in the 

 ciliary girdle. When in a normal, fully expanded state, this is 

 borne at the edge of a widely flaring portion of the posterior end 

 of the larva, PI. XVII, Fig. 3. At intervals along this expanded 

 part occur scalloped thickenings, the convex edges of which 

 are directed inward and anteriorward. About ten of these were 

 present on the specimen from which Fig. 3 was drawn. When 

 the animal is in a considerably contracted state, as, for example, 

 was the case with this particular specimen when first seen, the 

 intervals between these scallops stand out in prominent lobes, or 

 flaps, PI. XVII, Fig. 4. 



The course of the anterior, ciliated bands is unusual. Instead 

 of following the simple, typical meander found in such species 

 as T. mulleri and T. agassizi, or 'if being produced into distinct 

 tentacles as in T. grenacheri and its relatives, they have more 



*Spengel's plan of giving tornariae the names of the authors who first describe 

 them cannot be followed in this instance. We. however, come as near conformity 

 as possible and gladly name this one after Miss Marian Hubbard, who rendered 

 valuable assistance in the laboratory work connected with the preparation of this 

 paper. 



