BULLETIN OF THE UNITED STATES FISH COMMISSION. 155 



mesoblastic roof of the cardiac portion of the segmentation cavity lying 

 beneath the head ; at first there is no definite arrangement of the cells 

 destined to become the heart, but they seem to be spread out in a loose 

 mass between the hypoblast and the mesoblast at the point where the 

 heart will apx)ear. As soon as they have grown down and come into 

 contact with the hypoblast a circular space or cavity is formed in their 

 midst, which is the rudiment of the heart of the mackerel in its sim^jlest 

 l^ossible form. It is now nothing more than a wide circle of coarse cells 

 interposed between the mesoblast and hypoblast, so that one may look 

 through the lumen or opening in the ring either from above or below. 

 In the process of growth this ring of cells is drawn out into the x)rimitively 

 simple tubular heart, the hypoblastic or venous end being dragged for- 

 ward while the branchial or aortic end is directed backwards. Thanks 

 to the transparency of these embryos, every step of the process may be 

 seen just as I have described it. By the eighteenth hour the heart h, 

 Fig. 12, is fusiform and open at the venous end, and still bound to the 

 hypoblast, and now begins to contract slowly and at long intervals, al- 

 though there are still no blood corpuscles visible in the fluid held in its 

 cavity. The next change observed is from this i)oint onwards, when its 

 anterior end is bent to the left and finally opens backward, and it is 

 now clearly determined that the wide backwardly directed portion will 

 become the venus sinus and Cuvierian ducts ; the point where the bend 

 is made will become the ventricle and the other narrow end the bulbus 

 aortae. At no time, nor in any form, have I seen any evidence of the 

 origin of the cavity of the heart by the coalescence of two distinct spaces, 

 as described in the develoi^ment of other types of vertebrates. 



The embryo on the eve of hatching has a relatively shorter tail than 

 most other tji^es of true fishes, and when just hatched measiu-es a 

 little more than one-eleventh of an inch in total length. It usually 

 escapes head first from the egg, and manifests a singularly quiescent 

 disposition, but as it grows older and begins to right itself, as its 

 oil sphere becomes smaller, it will settle on the bottom of the vessel 

 in which it is confined, but if disturbed it will dart oft" and out of the way 

 with great quickness, and shows a disposition to avoid danger. The 

 yelk has diminished in bulk before the egg-membrane is ruptured, be- 

 cause the embryo fish has grown at its expense, and a considerable 

 quantity of i)rotoi)lasmic matter has doubtless been budding off from 

 it in consequence of the formation of free nuclei, which are found, in 

 other species at least, in the sui)erficial layers of the yelk just below 

 the embryo. The diminution of the bulk of the yelk is not due to the 

 development of the blood, which is not jet discoverable, nor will it 

 ai)i>ear until some time after the young fish has left the egg. 



The rudiments of the breast fins appear just before hatching as a 

 pair of deUcate rounded folds, which, have a horizontal direction at the 

 base, and which grow out on each side of the body in the vertical from 

 the oil sjihere. They may be regarded, therefore, as having very little 



