[87] EMBRYOGROPHY OF OSSEOUS FISHES. 541 



understand that I shall bere reproduce what Balfour has said of it iu 

 his Comparative Embryology, II, p. 519: 



•' The actual observations bearing on the origin of the vascular sys- 

 tem, using the term to inclnde the lymphatic system, are very scanty. 

 It seems probable, mainly it must be admitted on a priori grounds, 

 that vascular and lymphatic systems have originated from the conver- 

 sion of indefinite spaces, primitively situated in the general connective 

 tissue, into definite channels. It is quite certain that vascular systems 

 have arisen independently in many types ; a very striking case of the 

 kind being the development in certain parasitic Copepoda of a closed 

 system of vessels with a red non-corpusculated blood (E. Van Beneden, 

 Heider), not found in any other Crustacea. Parts of vascular systems 

 appear to have arisen iu some cases by a canalization of cells. 



" The blood systems may either be closed, or communicate with the 

 body cavity. In cases where the primitive body cavity is atrophied or 

 partially broken up into separate compartments {Insecta, Mollusca, 

 Discophora, etc.), a free communication between the vascular system 

 and the body cavity is usually present ; but in these cases the com- 

 munication is no doubt secondary. On the whole it would seem prob- 

 able that the vascular system has in most instances arisen independ- 

 ently of the body cavity, at least in types Avhere the body cavity is 

 I)resent in a well developed condition. As pointed out by the Hert- 

 wigs, a vascular system is always absent where there is not a con- 

 siderable development of connective tissue. 



" As to the ontogeny of the vascular channels there is still much to 

 be made out both iu vertebrates and invertebrates. 



" The smaller channels often arise by a canalization of cells. This 

 process has been satisfactorily studied by Lankester in the Leech,* 

 and may easily be observed in the blastoderm of the chick or in the 

 epiploon of a newly-born rabbit (Schiifer, Eauvier). In either case the 

 vessels arise from a network of cells, the superficial protoplasm and 

 part of the nuclei giving rise to the walls, and the blood corpuscles be- 

 ing derived either from nucleated masses set free within the vessels 

 (the chick), or from blood corpuscles directly differentiated iu the axes 

 of the cells (mammals). 



" Larger vessels would seem to be formed from solid cords of cells, 

 the central cells becoming converted into the corpuscles and the pe- 

 ripheral cells constituting the walls. This mode of formation has been 

 observed by myself in the case of the spider's heart, and by other 

 observers in other invertebrata. In the vertebrata a more or less similar 

 mode of formation appears to hold good for the larger vessels, but fur- 

 ther investigations are still required on this subject. Gotte finds that 

 in the frogs the larger vessels are formed as longitudinal spaces, and 

 that the walls are derived from the indifferent cells bounding these 

 spaces, which become flattened and united in to a continuous layer. 



Quart. Joiirn. Mic. Science, Vol. XX, 1880. 



