526 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [72] 



the process. I believe also that the segmentation cavity is a lymph 

 space, and that, since the first blood-corpuscles are borne in it, the evo- 

 lution of ])lasmine and fibrin may occur within it at an early stage and 

 aid in such a process as the outgrowth ol the tail, and thus indirectly 

 in the development of its lateral muscles. 



At the time of the closure of the blastoderm the number of muscle- 

 segments developed in different species is also subject to considerable 

 variation ; so marked is this in extreme cases that it is i)roper to call 

 attention to it in this connection. We find, for example, in many forms, 

 not more than eighteen to twenty muscle-segments developed on either 

 side of the body up to the time when the blastoderm closes over the 

 yelk. In exceptional cases, as many as seventy-five maj' be developed, 

 as we find in the instance of Tylosurus. This variation is doubtless due 

 to the influence of heredity, the embryos which have the most segments 

 at an early stage descending from adults which have a proportionally 

 large number of muscular segments developed, while those embryos 

 with but few are descended from parents with a less number. 



The inclusion of the yelk-sack, or what remains of it at a late stage 

 of development in young salmon, by the downgrowth of the ventral 

 ends of the muscular segments overlying the sides of the abdomen, is 

 a very interesting phenomenon. It recalls in some respects the pro- 

 cess of inclusion of the yelk by the blastoderm at a much earlier stage. 

 Unlike the latter, however, they do not coalesce at one point or come 

 together as a round pore of gradually lessening diameter like the rim 

 of the closing blastoderm, but they join on the median ventral line 

 from the isthmus back to the pre-anal fin-fold; the opening which re- 

 mains at this time between the ends of the down-growing episkeletal 

 muscle plates has the form of a very elongate median, ventral cleft. 

 The abdominal cavity in the young salmon is also relatively long in 

 contrast with that of the young cod, but in larval Clupeoids it is of 

 still greater relative length than in the salmon and proportionally 

 longer than in any other forms known to me of the same stage. The 

 downgrowth of the lateral muscle plates in all Teleostean types ai)pears 

 to take place in a somewhat similar manner to form the episkeletal mus- 

 cular stratum external to the ribs. 



20. — Devjclopment of the intkstiise and its appendages. 



The development of the intestine of the Telcostei or true fishes is 

 peculiar in a number of respects ; these are, first, its primitively solid 

 and depressed form, and secondly, the mode in which the oral end of it 

 appears to be developed from behind forwards, there being apparently no 

 clearly marked oral invagination of the epiblast or a stomoda^um; 

 thirdly, the mode of formation of the proctodteum or anus; fourthly, 

 the ai)i)earance of a lumen in it not by a process of invagination from 

 below or behind, but by a separation or retreat of its cells from its axis. 

 Like the intestine of other vertebrates it is develoijed from the true 



