[81] EMBRYOGEAPHY OF OSSEOUS FISHES. 535 



in the head region, and in the last along almost the whole length of 

 the ducts at the time of hatching. In Gamhusia the segmental tubules 

 form a complex, convoluted mass just behind and partly below the ear, 

 and is richly supplied with blood long before tbe young have absorbed 

 the yelk-sack or have been discharged from the ovarian follicles of the 

 parent. Along the segmental ducts or pronephric canals, behind this 

 point, there is no evidence of tubules whatever, yet in the adults of 0am- 

 husia we find the WolfBan body or kidney extending dorsally along the 

 whole length of the body cavity. No such complex head-kidney, as we 

 may call the structure found in the embryos of OamJnma, is developed 

 in the young of Salmo, even at the time of hatching, although segmental 

 tubes have already been formed. An examination of Alosa of the same 

 relative age shows that absolutely no segmental tubes or accessory glo- 

 meruli are developed. To what cause are we to assign this difference? 

 The cause is apparently a physiological one, and is probably not due 

 to any phylogenetic influences, except as these may be expressed in an 

 accelerated or retarded state of development of the systemic circulation. 

 Both in Gambusia and Salmo, of the stage of development here consid- 

 ered, the blood vascular system is already far advanced, while in Alosa 

 and Gadus there is still no circulation ; this seems to me in part at least 

 to offer an explanation of the great difierences observed in the develop- 

 ment of these organs in the embryos of the same age of different genera.. 

 As regards the local or general development of the mesonephric struct- 

 ures along a part or the whole of the segmental ducts, that difference 

 is of course probably to be ascribed to hereditary or phylogenetic, and 

 not to physiological influences. 



The value of the evidence regarding the opening of the glomeruli into 

 the body cavity will depend altogether upon what is meant by the latter 

 term. If it is held, as it is by me, that the body cavity of fish embryos 

 is the same as, or is at first continuous with, the segmentation cavity, 

 then the glomeruli, as far as I am able to interpret my sections, are shut 

 off from the body cavity. This is the view also which I should take of 

 the sections figured by Balfour, Zeigler, and CEllacher. The segmental 

 tubes are probably developed, like the glomeruli, from mesoblast, which 

 lies above the peritoneal layer and between it and the aortic and ven- 

 ous vascular tract. In sections through the pronephros of Alosa it has 

 appeared to me as if it opened anteriorly into the body cavity, but 

 I could see no evidence of a glomerulus; but this, it is to be remem- 

 bered, was in embryos which had not yet developed a circulation. The 

 peritoneal or splanchnopleural layer is well marked in embryo fishes, 

 but it does not usually extend far out over the yelk in early stages, so 

 that it is easy to see that it cannot include the yelk. The true hypo- 

 blast, after the development of the intestine in Alosa, seems to have 

 vanished as a discernible layer, so that the gut lies directly upon the 

 yelk, and is therefore bounded on either hand by the segmentation cav- 

 ity, which is naught else but the body cavity itself, which diminishes in 



