62 The ll'est American Scientist. 



two pseudopods; the dendritic form is not developed for some 

 time. The color granules of most of the cells are brownish- 

 yellow; those of comparatively few cells being black. The cells 

 are so large that they may readily be distinguished before any 

 color is deposited in them. The term chromatoblast may prop- 

 erly be restricted to them at this colorless stage. There are no 

 chromatoblasts distinguishable when the embryonic shield begins 

 to form; they first appear when the gastrula covers one-third or 

 one-half the yolk. They are proliferated info the segmentation 

 cavity from the (nibryo, where the latter joiiis the embryonic riyig 

 and Jrovi here they migrate to various portions of the ernbryo. 

 Just how they arise I cannot say, as means for sectioning are not 

 at hand. Further observations of living eggs may clear this 

 point. As the gastrula advances over the yolk they increase in 

 number very rapidly, but, as tar as I have observed, by the pro- 

 cess of proliferation only. 



Thev are somewhat angular in shape, the angles being some- 

 times prolonged into pseudopods. Their depth at this time is 

 equal to that of the segmentation cavity. In their migration the 

 pseudopods are usually foremost; amoeboid changes in shape 

 were not observed, the whole cell moving forward. Most of the 

 chromatoblasts are collected along the margin of the embryonic 

 shield. When the embryonic shield narrows and the embryo 

 becomes more definitely outlined the chromotophores rapidly mi- 

 grate inward and cover the embryo, while others collect about 

 the oil-globules and but few remain over the yolk. Shortly 

 after the closing of the blastopore in Scieana the oil-globules 

 unite and the eggs ot this species can then be distinguished from 

 those of Hypsopsetta only by the very slight difil'erence in size. 



The chromatoblasts cannot be confounded with any other 

 elements. They are approached in structure and position by 

 the cells of the epiblast when about to divide. Such cells rise 

 out from among the surrounding cells and project into the seg- 

 mentation cavity, divide, and again sink to the ordinary level. 



As chromatoblasts cannot be observed continuously from their 

 origin to the chromatophore stage, should this process extend 

 over several days or weeks, it is not possible to say positively 

 whether the color-cells of the genera Oligocottus, Isesthes, etc., 

 have a similar origin. The fact that their genesis in so widely 

 different genera as Hypsopsetta and Sciaena is identical, would 

 indicate that the color-cells in all fishes are developed in a similar 

 manner. Carl H. Eigenmann. 



* Preliiniiuiry Notice. 



CONTRIBUTIONS TO WEST AMERICAN BOTANY. 



In this series of papers it is proposed to present in a connected 

 form, such botanical observations as may be made by the writer, 

 that may seem to contribute to a fuller knowledge of our West 

 American fiora. It is hoped that these may become worthy of the 



