﻿oQb -ENDEAVOUR" SCIENTIFIC REST'LTS. 



varies very much in the different specimens examined, for here 

 there may be either a median pair of tentacles or a median 

 unpaired one. I have to thank Miss Ethel Summons, Mel- 

 bourne University, for first drawing m\ attention to the paired 

 arrangement. In PI. xxxvii., figs. 8 to 15, I have represented 

 the bases of the oral cartilages in the mid-ventral portion of 

 the hood of eight different' specimens, and it will be seen by 

 them that there is a great amount of \ariation here. PI. 

 xxxvii., figs. 8, 9, 10, II, ij show no median unpaired ten- 

 tacle, but a distinct median pair, while in PI. xxxvii., figs. 13, 

 14, 15 there is a median unpaired one. In PI. xxxvii., fig. ib, 

 the margin of the oral hood with the cirri cartilages is repre- 

 sented, and it will be seen that the basal portions of the 

 cartilages of the median pair are smaller than the others and 

 are inclined at a different angle to the margin of the hood, and 

 so are clearly distinguishable. The two \entral cirri repre- 

 sented in PI. xxxvii., fig. u, are not so ciearl}- paired as in 

 the others. 



In referring to the literature on the subject I can find no- 

 thing very definite. In 1895, Miss J. W. Kirkaldy,! giving 

 the characteristic features of the Family Branchiostomida?, 

 says: — "A prjeoral muscular hood is dexeloped in front of the 

 mouth, provided along its circular margin with numerous 

 tentacles, supported by a cartilaginoid skeleton ; there is one 

 median unpaired tentacle in the median \ entral line, and from 

 ten to twenty (according to age and species) on either side." 



In 1876, Peters2 described Epii::<uiiclitliys citlfclhtni as 

 having ten to twehe pairs of oral cirri, and figuring the under 

 surface of the anterior end draws the cirri show ing clearl} that 

 there is no median cirrus, but he does not state definitely the 

 presence of a median ventral pair. Miss Kirkaldy, ho\\e\er, 

 in her paper, describes this species as having forty-one to 

 fortv-three oral cirri, thus giving it a median tentacle. 



Then, again, Mr. j. D. F. Gilchrist^ gives the number of 

 oral cirri in Branchiosionia capensc as thirty-six in all. It, 

 as is generally taken to be the case, the cirri are paired on the 

 two sides of the hood (with the exception of the median \ en- 

 tral one), i.e., if the number of cirri is the same on the two 

 sides, this would not allow of a median cirrus being present 

 in this species. It is thus evident there is some uncertainty as 

 to the exact arrangement of the cirri and that a structure 

 previously described as characteristic of the Family Branch- 

 iostomidge is actually subject to variation within the limits of 

 a single species. 



1 J. W. Kirkaldy— Quart. Jour. Micro. Sci., xxxvii. 'n.s.). 1895. p. 309. 



2 Peters— Monat. K. Preuss. Al?ad., 1876. 



3 Gilc-hrist— Marine Investigation in South Africa, ii.. 1904. p. 111. 



