PEROPIIORA ANNECTENS. 55 



of the test is produced as a secretion of the mesodermal 

 cells imbedded in it. It seems rather to be the product 

 of the ectodermal cells by which it is lined: and this 

 agrees with the more usual view of its origin. In many 

 places, both in the stolons and in the developing zooids, 

 the cells of the ectodermal layer have such a form as is 

 shown in fig. 39, pi. iii. I believe this to be due to the 

 fact that the cellulose substance of the test is here being 

 formed. The processes are probably similar to the ones 

 described by Salensk}' ('91? P- 14)5 in Pyrosoma, as like- 

 wise projecting from the ectoderm cells into the test. 

 This author also regards the processes as having to do 

 with the formation of the cellulose substance. He says, 

 however, that they are found only at an early stage in the 

 development of the individual. 



3. THE MUSCULATURE. 



The musculature of the mantle is contined to the 

 anterior end of the body, as in other species of the genus. 

 As seen in the whole animal, when examined as a trans- 

 parent object, the longitudinal fibres are most conspicuous 

 as they run backward, separated by wide and quite regular 

 intervals, into a region where the test is particularly thin 

 and transparent. (Fig. 3, r. m., pi. i.) As seen by this 

 figure, they are radially arranged, each bundle of fibres 

 growing gradually smaller as it passes backward, finally 

 disappearing entirely, usually before the second transverse 

 branchial vessel is reached. A small fragment of the 

 mantle from the branchial siphon is shown in fig. 18, 

 pi. ii. The specimen is seen upon its external surface, 

 consequently the circular fibres are generally situated 

 internally to the longitudinal fibres. This, however, is 

 not always the case, as for example, the radial bundle 

 r. Ill/ passes under some of the circular bundles and 



