A TAXONOMIC STUDY ON PYROSOMA — METCALF AND HOPKINS. 261 



SPECIATION. 



As noted in the introduction to this paper, one of the features of 

 Pyrosoma, of chief interest, is the remarkable intergradation shown 

 between its different forms, and this is the more impressive when 

 one compares this family with the family Salpidae, whose species are 

 sharply distinct from one another. The conditions among the 

 Pyrosomas suggest hybridization as a factor cooperating with muta- 

 tion to produce the results observed. In the atlanticum group, at 

 least, hybridization seems probably to be continuing to-day. 



The differences between species are far greater among the Salpidae 

 than among the Pyrosomidae. Lack of physiological isolation within 

 the latter group has probably aided to prevent extreme divergence, 

 but the inherent tendencies to divergence were doubtless also less. 

 Pyrosoma is a remarkably stable form, its several species and sub- 

 species differing from one another but slightly, and this chiefly in 

 size of colony and in the character of the test processes, in the relative 

 proportions of the series of three respiratory chambers in the zooids, 

 and in the number of stigmata and of branchial bars. 



INDIVIDUALITY AND FORM CONTROL. 



Another keenly interesting feature in Pyrosoma is the way it pre- 

 sents the universal problems of form control and of individuality, 

 two very closely related conceptions. Pyrosoma, like all other 

 colonial organisms, has three grades of individuality, that of the cell, 

 that of the zooid and that of the colony, but it is peculiarly interesting 

 in that the individuality of the cell is not only subordinated to that 

 of the zooid, but is also, in the case of some cells, directly subordi- 

 nated to the colony as a whole, without reference to the zooids. This 

 feature is best seen in connection with the form and sculpturing of 

 the colonial test. 



Note first a feature characterizing the oral test processes in Pyro- 

 soma atlanticum atlanticum (fig. 30, pi. 26, and fig. 33, pi. 28), or in P. 

 ovatum (fig. 29, pi. 25). Each elongated test process is obliquely 

 truncated distally, the area of truncation being smooth and nearly 

 Hat, and containing the oral aperture. Outside the truncate surface, 

 the tip of the oral test process bears numerous minute denticles, at 

 the base of each of which lies a tost cell, which doubtless either secretes 

 the test material of which the denticle is composed, or so influences 

 its arrangement that it takes the form of a denticle. Those test colls 

 are mesenchyme cells, which have wandored outside the zooids in 

 which they arose, have migrated to a distance, and have oach taken 

 up a definite position in the test at a distance from the zooids, there 

 to form or control the formation of a denticle, a minute protruding 

 bit of test substance. They take their position, not on the area of 



