44 



THE VOYAGE OF H.M.S, CHALLENGER. 



pass to the gill-slits by a very slight alteration of the direction of these currents, since 

 the gill-slits open ventro-laterally from the first division of the alimentary canal. Solid 

 particles conveyed by the current could presumably be readily retained in the alimentary 

 canal, whilst the water itself would naturally pour out by the gill-slits, which would be 

 protected from the afferent current due to the tentacles by means of the operculum. It 

 thus seems possible to conceive of a method in which the gill-slits may be of the greatest 

 importance in connection with the straining of the water from the food-particles. Bateson 

 has been led to assume that gill-slits are structures which have been developed within 

 the Chordata, and it is suggestive that in Cephalodiscus, owing to the rich development 

 of the tentacles, we appear to have a cause capable of inducing the evolution of perfora- 



Fio. 4.— Transverse section through the collar and prohoscig of an adult Cephalodiscus, passing in front of the mouth : n.s., 

 nervous system; p., proboscis; l, two of the lophophoral arms cut horizontally; jr., grooves continuous with the 

 grooves of the anterior lophophoral arms ; ft. c. ', body-cavity of proboscis ; 6. c", body-cavity of collar; «cA.,notochord ; 

 D., dorsal, and V., ventral surface. 



tions leading from the alimentary canal to the exterior. It may be perhaps profitable to 

 consider whether gill-slits may not possibly have originated in this manner, and whether 

 the arrangement found in Balanoglossus and other types may not have been developed 

 from a CephalodiscusASk% condition b}^ the metameric repetition of the slits, accompanied 

 by a change in their function. 



The preceding remarks relating to the supposed affinity of Cephalodiscus to Balano- 

 glossus may with advantage be summed up by a statement of the various features which 

 are common to the two genera. 



1. Division of the body into proboscis, collar and trunk, this division being specially 

 obvious, in Cephalodiscus, in the young bud. 



