950 DR. W. G. MBEWOOD ON THE [NoV. 28, 



The efferent branchial vessels of Chanos have been described 

 and figured by Hjrtl (12. pi. i. fig. 1). The first and second 

 vessels open into the circulas cephalicus, the third some distance 

 farther back into the median aorta, vi'hile the vessels from the 

 fourth gills, after uniting with those from the epibranchial organs, 

 unite with one another and open into the aorta at a point as far 

 behind the opening of the third branchial vessel as the latter is 

 behind the circulus cephalicus. The genus thus falls into group B, 

 but, owing to the exceptional disposition of the fourth pair 

 of vessels, it cannot be included in any of the three subgroups 

 recognized. 



FURTHEE EeMAEKS ON GrROUP C. 



The circulus cephalicus is narrotv in front in Lahrax (fig. 17), 

 but it is broad in the ISiluroids Liocassis (fig. 18), Callichthys 

 (fig. 19), and Clarias (fig. 20). In Lahra.v the cceliaco-mesenteric 

 artery arises from the aorta immediately behind the fourth 

 efferent branchial. It has the same relations in Clarias, but is 

 somewhat more posterior in Callichthys, and considerably so in 

 Liocassis. The subclavian arteries arise behind the coeliaco-mesen- 

 teric artery in Lahrax, Clarias, and Callichthys, but close behind 

 the fourth branchial vessel in Liocassis. Although Clarias is 

 introduced into the subgroup Ca, it really occupies an intermediate 

 position between C a and C b, since there is a very short common 

 trunk on each side between the last two branchial vessels and the 

 aorta. The genus Heterotis I have not been able to examine, but 

 it is tolerably certain, from the description given by Hyrtl (9. p, 87), 

 that it should be included in the subgroup Ca. 



In Mullus and Gorvina the disposition of the coeliaco-mesenteric 

 and subclavian arteries is as in Ptrca (fig. 21), but the circulus 

 cepljalicus is broader in front. The circulus has the form of a 

 regular heptagon in Mullvs, while in Corvina it is pear-shaped, 

 the broad end being anterior. In Equula it is oval in shape and 

 longer than broad ; but otherwise the relations of the parts are as 

 in Ptrca. With regard to Perca itself, Hatchett Jackson, in his 

 edition of Eolleston's 'Forms of Animal Life '(21. p. 88), states 

 that the " coeliaco-mesenteric artery * * * springs from the right 

 epibranchial artery before it fuses with its fellow." This does not 

 accord with my own observations. Hyrtl in his figure of Lucio- 

 ■perca (7. pi. i. fig. 1) shows the common trunks of the third and 

 fourth branchial vessels entering the aorta at some distance from 

 the posterior angle of the circulus. The transverse commissure, 

 also, between the anterior carotids (" die vorderen oder kleinen 

 Kopfarterien ") is situated farther forward than in Perca. 



In Zeus (fig. 22) the circulus is broader, and the coeliaco-mesen- 

 teric artery arises, not from the aorta, but from the common 

 trunk of the last two branchial vessels of the right side. Zoarces 

 resembles Anarrhichas (fig. 23) in the oval shape of the circulus 

 and in the narrow anterior prolongation of the latter, but the 



