28 DEVELOPMENT OF LYMPHATIC SYSTEM, FISHES 



wise clear lymph space and, as is the case of the subocular sacs, 

 are undoubtedly related to the mesenchj^mal strands which 

 formerly bridged the space. 



I have been unable to determine the presence of a definite 

 endothelium in the earliest anlagen of the subocular lymph sacs. 

 The cells which form their walls cannot be distinguished from 

 those of the surrounding mesenchyme. As soon as the subocu- 

 lar lymph sacs assume the character of definite collecting centers 

 for lymph, the cells which form their walls begin to arrange 

 themselves in the form of a definite endothelium. Such a con- 

 dition signifies that certain cells of the embryo are undergoing 

 a differentiation to form the definite lymphatic endothelium, 

 otherwise it is difficult to see why transitional stages should be 

 found. 



Exactly the same transitional stages from mesenchyme to en- 

 dothelium can be observed in connection with the establishment 

 of the definite endothelium of the veins. At an early stage of 

 development the cells which form the walls of venous channels 

 cannot be distinguished from those of the surrounding mesen- 

 chyme and it is only later that they become spindle-shaped 

 and assume the typical endothelial form. 



Are the early anlagen of the subocular lymph sacs connected 

 with the endothelium of the veins by solid cords of cells which 

 might indicate that they are outgrowths from the veins? 



I have also been unable to determine the presence of solid cords 

 of cells which might extend between the endothelium of the veins 

 and the anlagen of the subocular lymph sacs. The only veins 

 with which such solid cords of cells could posibly be connected 

 are the precardinal (jugular) and the infraorbital veins. The 

 precardinal veins (6, figs. 19 and 20) lie dorsally and widely sepa- 

 rated from the sacs. The infraorbital veins (23, figs. 1 and 2) 

 which are tributaries of the precardinal, lie at first anterior to the 

 anlagen of the subocular sacs, and in the embryos from which 

 figures 19 and 20 are taken, are situated about sixty and seventy 

 micra, respectively, anterior to these anlagen (compare 1 and 23 

 in figs. 1 and 2). Figure 25 is a photomicrograph of a section 

 taken through a fourteen-day steelhead trout embryo at the level 



