22 AUGUSTA ÄRNBÄCK-CHRISTIE-LINDE, NORTHERN AND ARCTIC INVERTEBRATES. 



composed of follicle, oviduct and one egg, thus it is of exactly the same structure 

 as the ovary of the Botryllids described. From the investigations of Iiile 1 it would 

 seem that there are ovaries of analogous structure in the Salpidce also. A true pla- 

 centa has been described and ring-shaped folds develop to attach the embryos, though 

 of a far more complicated structure than the corresponding organ in Botryllus schlosseri. 



The relationship betvveen the Botryllidce and Polyzoidce has often been alluded 

 to in literature without any change having resulted as to their accepted systematical 

 position to each other. Owing to the recent subordination of the Polyzoidce under 

 the Styelidce, the gap between them, from a classificatory point of vievv, seems to 

 have become still wider than it was. There are, however, good reasons justifying a 

 different conception as to their relationship to each other. From a comparison be- 

 tween them we may discover the following points: 



The branchial sac of the Botryllids agrees in all essential points with that of 

 certain Polyzoids, viz. Kiikenihalia borealis and its ally Chorizocarpa. 2 In both groups 

 it is characterized by the absence of folds and the presence of a few longitudinal 

 vessels, in the above-mentioned Polyzoids from three to five in number, in the Bo- 

 tryllids described three. In this connection it may be pointed out that in no Bo- 

 iryllid species is a larger number known for certain as yet. The statement that in 

 Bolrylloides gregalis, collected off South Africa and described by Sluiter 3 , there are 

 five longitudinal vessels and that in Bolrylloides chazaliei, a West Indian form de- 

 scribed by the same author 4 , there are four on each side, is, I think, based upon 

 some mistake. On re-examining the two species kindly placed at my disposal by 

 Sluiter, I have not observed more than three longitudinal vessels on each side in 

 either species. 



The alimentary canal affords another point of agreement. In both groups the 

 stomach is short and well marked-off from the intestine; it is provided with numer- 

 ous longitudinal folds; the pyloric coecum as a rule is well developed, the intestine 

 forms a short, rounded loop, the rectum is short, and the margin of the anus is 

 generally plain. 



As to the generative organs several characteristic features are common to both 

 groups and further investigations will probably further add to the number of them. 



If we compare the male organs of Chorizocarpa guttata with those of the Bo- 

 tryllids we find that, in the main, they are characterized by the same shape and 

 structure in both the gland and the vas deferens. 



Judging from what we know about Kukenthalia borealis, the occurrence of one- 

 egg ovaries, a peculiar feature in the Botryllidce, is probable even in the Polyzoidce, 

 at least in certain genera. A satisfactory comparison cannot, however, be made in 

 respect to the female organs, as the investigations are not yet completed. Want of 



1 Ihle, J. E. W., l)as Tierreich. Lief. 32. Desmomyaria. Berlin 1912. 



2 Michaelsen, W., Revision der compositen Styeliden öder Polyzoincn. Mit. Mus. Hamburg. Jabrg. 21. 

 Hamburg 1904. 



3 Smiter, C. Ph., Tunicaten von Siid-Afrika. Zool. Jabrb. Syst. Bd. 11. Jeua 1897. 



4 Sluiteu, C. Pli., Tuniciers recueillis en 1896 par la Cbazalie dans la mer des Antilles. Mém. Soc. 

 Zool. France. Vol. 11. Paris 1898. 



