516 REPORT OF COMMISSIONEK OF FISH AND FISHEKIES. [92] 



agaiu breaking up into smaller vessels in the liepatic tissue, from which 

 the blood again emerges to be conveyed in larger capillaries over the 

 yelk-sack, and which pass outwards, downwards, and forwards, to empty 

 into the great median vitelline vein on the ventral face of the yelk, a 

 little to the left side of the middle line. The anterior and posterior 

 cardinals together with the great vitelline vein empty their blood into 

 the venous sinus, from whence it is passed into the heart, and from thence 

 through the branchial vessels is sent through the carotids and aorta, 

 the latter of which is supi)lied by the combined currents from the four 

 hinder branchial vessels, which converge and meet in a common aortic 

 trunk below the medulla oblongata. The subclavian artery of the pec- 

 toral arises from the vicinity where the branchial vessels unite into the 

 aortic trunk. The origin and course of the submaxillary and cephalic 

 vessels I have not made out. Supraocular and postcerebellar veins 

 l^ass backwards on the head to empty into the jugulars. 



The somatic capillaries are somewhat interesting in respect to their 

 arrangement. They are given off from the aorta and pass outwards on 

 either side through the muscular septa between the muscular somites 

 on the middle line of the side, all of them traversing the common sep- 

 tum which divides the dorsolateral from the ventro lateral portions of 

 the muscular i^lates. While these vessels are of a capillary character 

 in the embryo, they become the segmental arteries and veins of the 

 adult. The course of the blood current in them is not in the same 

 direction in all of them, however ; in some it is afferent and in others 

 efferent in direction, so that it would appear that some of the segmental 

 vessels were really venous and others arterial. After reaching the sur- 

 face the arterial segmental vessels divide dorsally and ventrally into 

 branches which follow the courses of the intermuscular sejjta to pass 

 inwards at the dorsal and ventral borders to unite with the cardinal or 

 caudal veins. The venous segmental vessels are supplied from two 

 vessels which have exactly the same course as the intersegmental ca- 

 pillariesof the arterial segmental vessels, but the blood-flow is outwards. 

 They bend over the upper and lower edges of the muscle plates, follow 

 the septa, and at the middle line of the side, at the level of the horizon- 

 tal septum, between the dorsolateral and ventro-lateral plates, converge 

 into an incurrent segmental vein. These intersegmental veins and 

 arteries do not alternate regularly; there may be two arterial vessels 

 in succession followed by a vein between the next two following seg- 

 ments. The dorso- ventral intersegmental capillary loops convey the 

 blood from the aorta and to the great veins, so that in the case of a 

 true intersegmental vessel it may have either a single or a double origin 

 from the aorta, according as the outgoing vessel passes directly to the 

 surface at the middle of the side or by way of dorsal and ventral arte- 

 rial loops. The mode in which these vessels are channeled out in the 

 body I have not been able to make out. 



In Coregonus the vitelline vascular system is not *so complex, but, 



