cytologic detail of the immature thrombocyte nu- 

 cleus was not obsei-ved by Sugiyama (1926), 

 probably ])ecause Wright's stain was used. 



Beyond the blast stage, thrombocytes can be 

 identified easily; thus even in one field ( fig. 225) 

 two older stages are visible. Cells 24 and 26 

 are probably at about the same level of differen- 

 tiation, and cells of this size do not tend to clump 

 together as readily as do the smaller cells (27). 

 In the clumping process, cytoplasmic IdeJjs are 

 thrown off. A duplicate slide, made from the 

 same eml)ryo that was used for figure 22.5, was 

 stained with Ralph's modification of the benzi- 

 dine test. Two cells ( figs. 277 and 278 ) demon- 

 strate that hemoglol)in is absent from the throm- 

 bocyte even at an early stage of differentiation. 

 Figure 277 is probably equivalent to cell 26 in 

 figure 225; figure 278 is smaller and the nucleus 

 is not visible. Dividing throndjocytes are some- 

 times found. Unlike the erythrocytes which 

 maintain a compact cytoplasm, the thrombocytes 

 even during mitosis (fig. 279) can be stimulated 

 to give off protoplasmic fragments. 



Spreading and drying the cells is not always 

 rapid enough to prevent degeneration of the 

 thrombocytes. Many of the cells illustrated in 

 figures 276-295 are not typical of what one so 

 often finds on the slide, since the ones selected 

 were tliose least degenerated. A degenerated 

 throndjocyte looks like a lymphocyte. Not 

 every slide made from a series of embryos is 

 equally productive of early stages, but the em- 

 bryo that contril)Uted figures 280 and 286 had 

 an abundance of them. 



A typical throml)ol)last at 4 days of incuiiation 

 is shown in figure 280. It has a narrow rim of 

 cytoplasm that stains an intense dark blue or deep 

 violet. Within are large mitochondrial spaces. 

 The reticulum at the surface of the nucleus in fig- 

 ure 280 appears more clearly defined than that 

 in the nucleus of the amoeboid thromboblast in 

 figure 276. Associated with this more open net- 

 work is a vague suggestion of a nucleolus. 



Failureof the nuclear surface to stain is a com- 

 mon feature of young cells with large nuclei, and 

 three degrees of incomplete staining are shown 

 in figures 281-283. These duplicate the reac- 

 tion that occurred in young erythrocytes (figs. 

 2.59-261) in that the nucleolus is more readily 

 visil)le than in the well-stained cell. In these 

 three examples, deficiency in nuclear affinities 

 does not influence the (fuality of cytoplasmic 



staining: full differentiation of abundant mito- 

 chondrial spaces is attained, especially in figures 

 283 and 284. 



Shifting of the nucleus to an eccentric position 

 may occur in throml)oblasts (figs. 283-285) but, 

 as in the erythrocytes, the nucleus tends to remain 

 in the center of the cell, as in figures 280 and 

 286, and in the four small thrombocytes in fig- 

 ure 225, i.e. 27. By 4 days of incubation a wide 

 range in stages of development has appeared; the 

 cells form a series of decreasing size and increas- 

 ing cytoplasmic differentiation (figs. 286-292) ; 

 yet the impression is gained that this differentia- 

 tion process does not exactly parallel that which 

 takes place at older embryonic ages and after 

 hatching. Like the primary generation of eryth- 

 roblasts, this primary generation of thrombo- 

 cytes seldom produces an oval cell. Later, how- 

 ever, cells of this shape can be found frequently 

 if the preparation is made quickly enough. A 

 few cells with oval shape have been seen at 4 days 

 ( fig. 287) and again once or twice at 5 days, but 

 these cases are too rare to be called typical. 

 Sugiyama (1926) pictured a throml)Ocyte of oval 

 shape, taken from an emlnyo incubated only 2 

 days. We have never seen a thrombocyte having 

 this shape in an endn-yo this young. 



The first step in the differentiation process fol- 

 lowing the blast stage is the loss of staining af- 

 finity by the cytoplasm (figs. 286 and 288) and 

 the next is extensive vacuolization (figs. 290 and 

 291 ) . In figure 290 a single magenta granule 

 appears to be present outside the nuclear margin, 

 and in figure 291 such jjodies are numerous. 

 They are not the specific granules of the definitive 

 thrond)ocyte lint resemble extruded cliromatin 

 particles more than anything else. The delicate, 

 fine pink granules that are characteristic of de- 

 finitive thromliocytes never appear in the first 

 generation of thrombocytes; the two bodies shown 

 in figure 293 from an embryo incubated 5 days 

 21 hours are about the size of definitive granules 

 but belong to the chromidial type. 



Thromljoblasts soon disappear from the cir- 

 culating blood of the embryo, after which only 

 large, medium, and small embryo thrombocytes 

 are seen. By the ninth day the occurrence of 

 specific granules is fairly common. The gran- 

 ules may appear as early as the eighth day ( fig. 

 294) but many of the cells still do not show them 

 until later in development. They are present in 

 those cells selected for illustration chiefly be- 



131 



