10 ARKIV FÖR ZOOLOGI. BAND 4. N:0 *2o. 



branchial vessels; and seoondly, bj^ the progressive suppression 

 in length of the median aorta, bringing about an approxi- 

 mation of the dorsal or proximal ends of the last three effe- 

 rent branchial vessels on each side.» This last concentration 

 of the entrances of the efferent vessels leads sometimes to a 

 confluence of the third and foiirth of those, and as I have 

 found the condition to be in Spheroides, even in the case of 

 the second and third. According to the different degrees of 

 specialization as regards this point, Ridewood has classified 

 the fishes he has examined into several groups and subgroups, 

 a classification that separates forms which in all probability 

 are closely related or brings together rather differing types. 

 Ridewood points out himself, that this classification must 

 be looked upon merely as an artificial sheme, and finishes 

 his paper with the foUovving remark (p. 954): »It does not 

 appear, therefore, that the facts revealed by the present 

 inquiry can be applied with any probability of success to the 

 interpretation of the affinities of the family and larger groups 

 of Teleosts.» In this point he is, without doubt, quite right. 

 But I think his researches are still of a phylogenetical value 

 as far as it concerns the arrangement within a group. That 

 the evolution of the circulus and the arrangement of the 

 efferent arteries has proceeded in the manner in which Ride- 

 wood has stated, seems to me quite certain, but I think 

 that this specialization has proceeded indej)endently within 

 different groups of Teleostean fishes. The conditions in the 

 Plectognaths support this view strongly. What has caused 

 this specialization, Ridewood was not able to find out. He 

 did not find any relation to the shape of the head, nor to 

 the development of the epipharyngeal dentition, as it occurred 

 to him at the commencement of his inquiry. In the Plecto- 

 gnaths I have examined, I have not made any observations 

 as regards this problem. Perhaps a study of the correspond- 

 ing skeletal parts or of the development of the vessels may 

 bring forth facts that to some extent will explain these con- 

 ditions. In the above-mentioned paper Ridewood has de- 

 scribed the circulus cephalicus and the efferent branchial 

 arteries in three Plectognaths, viz. Balistes acideatus, Or- 

 thagoriscus truncatus and Tetrodon palembangensis. I have 

 examined myself the foUowing species: Balistes vetida, Lac- 

 tophrys trigornis, Diodon holacanthus and Spheroides testudineus. 



I 



