vesicle probably originates from the mesoderm, while that of the tuni- 

 cate undoubtedly in most cases arises from the endoderm, would have 

 been ample ground for rejecting absolutely a suggestion of homology. 

 To-day, however, even from a general point of view and without re- 

 gard to evidence touching the particular instance, most zoologists 

 would be less sure of the fatality of such a difficulty. But in the light 

 of the information produced by Morgan 2 , that the body cavities of 

 Tornaria "may arise as enteric diverticula, as endodermal prolifera- 

 tions, or even arise from mesenchymatous beginnings", and the farther 

 information furnished by Lefèvre 3 that in the bud of Perophora vi- 

 ridis the pericardio-cardie vesicle arises from "free amoeboid cells of 

 the blood"; while the same structure certainly arises as an out-pocket- 

 ing from the wall of the endoderm in the embryos and buds of various 

 other tunicates (Seeliger, van Beneden et Julin, Willey, Hjort, 

 Ritter etc.), the difficulty of origin ceases to be very formidable. 



The consideration which at present seems to me to weigh heaviest 

 in favor of the homology of the two hearts is not so much their simi- 

 larity in structure as the uniqueness of the type. It is a type 

 of heart that is absolutely without counterpart elsewhere in the animal 

 kingdom, and when one reflects on what it really is, functionally, he 

 finds it difficult enough to comprehend how it could have arisen once 

 to say nothing of its having arisen anew twice. 



xlccepting the now pretty well established hypothesis that the 

 blood vascular system of the metazoa is, in large part, a closed off por- 

 tion of the blastocoel, we may quite easily understand how that, as the 

 spaces closed in and narrowed down more and more to produce vessels, 

 the walls might become contractile in places, and thus initiate a heart of 

 the annelid or vertebrate type. Such formation would be an easy, and 

 a natural process, and one might suppose it would occur over and over 

 again in different groups of animals, as it undoubtedly has. When, 

 however, we note the steps that must have been taken to produce a heart 

 of the style now under consideration, and then try to imagine what 

 influences could have caused them, we find much difficulty. Observe 

 what we have: 1) A wholly closed vesicle with a capacious cavity. 

 How account for the cavity? 2) This vesicle primarily wholly free in 

 the blastocoel cavity. If it arose from the endoderm why dit it become 

 severed thereform, or of what use could it have been as a heart with no 

 connection with blood vessels? 3) An invagination of one side of 

 this vesicle to produce a second vesicle inside the first. What in- 



2 The Development of Balanoglossus. Joum. of Morph. 1894. Vol. IX. 



3 Budding in Perophora. Journ. of Morph. 1898. Vol. XIV. 



