Ahnliche Bilder lassen mich annehmen, dass die 

 sich befreienden Knorpelzellen an der Ausbil- 

 dung des Knochenmarkstiomas aktiv teilnelimen 

 koiinen. Mit Sicherheit kann man bloss ihre un- 

 mittelbare Teibiahme an der Blutbildung ver- 

 neinen. Es ist aber sehr schwer, etwas be- 

 stimmtes iiber ihr weitei-es Schicksal auszusagen, 

 weil sie ja wenigstens zum Teil, wie gesagt, all- 

 malilich alle histologischen Merkmale der umbe- 

 benden Mesenchymzellen annehmen und infol- 

 gedessen nicht mehr erkannt werden konnen." ^ 



Granulocytes in all stages of development from 

 granuloblast to mature heterophil are found in 

 figures 319 and 320. The developmental stages 

 will not be described here since they are to be 

 discussed fully under adult bone marrow, but one 

 characteristic noted at this time is the tendency of 

 heterophils, whether they be immature or mature, 

 to clump with other cells. Three heterophils lie 

 within marginal depressions of the osteoclast, 

 cell 10 of figure 319, and in figure 320 cells of 

 various types, including heterophils, have 

 clumped together. Developmental stages of 

 erythrocytes do not clump. Fennel (1947) re- 

 ports (p. 237) that "Giant . . . cells frequently 

 give rise to one or more granulocytes by the pro- 

 duction of cellular blebs. Such blebs ultimately 

 pulled away from the surface and became free. 

 Giant cells under other conditions fragmented 

 to form thrombocytelike cells." The cellular de- 

 tails shown in his drawings of this process taken 

 from vital-stained preparations are not sufficient 

 to permit determination of whether his giant cell 

 was an osteoclast, a macrophage, or a clump of 

 cells. Nothing has been observed in this study 

 to indicate that any type of giant cell ever pro- 

 duces thrombocytes or granulocytes. 



The marrow, during embryonic life, is in- 

 volved primarily in the production of granulo- 

 cytes and erjrthroc5^es, and although probably 

 more erythrocytes than granulocytes are actually 

 produced, it appears to be the other way around 



' Translation : The cartilage cells released from the capsule 

 mingle in the most intimate way with the elements of the 

 young mesenchyme that are penetrating into the same terri- 

 tory. As they gradually get farther from the cartilaginous 

 boundary they seem to stretch out and to form outrunners. 

 Similar pictures lead me to assume that the cartilage cells that 

 are freeing themselves can take an active part in the formation 

 of the bone marrow stroma. One can with surety only deny 

 their direct participation in hematopoiesis. But it is very hard 

 to state anything certain about their future destiny, since, as 

 has been said, they gradually take on. at least to some extent, 

 all the histological earmarks of the surrounding mesenchyme 

 cells, and hence cannot be recognized any longer. 



because the eiythrocytes are discharged into the 

 circulating blood and the granulocytes are held 

 in depots until after hatching. A study of the 

 whole matter of relative production rates in vari- 

 ous hematopoietic organs should be carried out. 

 It should include a study of the yolk sac. Hema- 

 topoiesis in the yolk sac has been omitted from 

 this study for two reasons — first, it does not lend 

 itself to the smear method and, second, it has been 

 covered by the extensive writings of Dantschakoff 

 (1908b) on sectioned material and of Sabin 

 (1920) on living preparations. The importance 

 of using the smear method for a study of erythro- 

 poiesis and granulopoiesis in the case of the pig- 

 eon has been emphasized by McDonald (1939), 

 who says (p. 293), ". . . The chief advantage 

 of the imprint method is that it brings out the finer 

 structural features, especially those of the nu- 

 cleus, which are so important in critical studies 

 of immature cells. . . ." 



Granuloblasts are much more abundant in the 

 bone marrow before hatching (cells 1-6, fig. 

 321) than after. Six of them are shown in one 

 field and they occur in the spleen at this age (fig. 

 330) in the same high concentration as in bone 

 marrow. Actually cells 2 and 4 are the least 

 differentiated; the nucleus lies in the center of the 

 cell and the rim of cytoplasm is narrow and stains 

 intensely blue. The remaining cells show defi- 

 nite changes leading toward the metagranulo- 

 blast — the nuclei have a slightly eccentric posi- 

 tion, some of them show points of chromatin con- 

 densation, and the cytoplasm is partially broken 

 up by mitochondrial rods. At the metagranulo- 

 blast stage, which is not represented on this plate, 

 the cytosome shows vacuoles and inclusions 

 by which a reasonably accurate guess can 

 be made as to the type of granulocyte the cell 

 will be when it is mature. This is not the case 

 in the blast forms shown in cells 1 to 6. 



Cell 7 of figure 321 is classified as a heterophil 

 mesomyelocyte but actually it has barely passed 

 the promyelocyte stage. The magenta rings, the 

 indefinite boundary of the nucleus, and the pale- 

 orange precursor bodies are all present in cell 

 7, but in addition there are a few darkly stained 

 orange bodies. Wlien they have elongated, the 

 orange bodies will be the definitive rods. Their 

 presence is the basis for calling the cell a meso- 

 myelocyte. Cell S is also a mesomyelocyte but 

 it is somewhat older than cell 7. All the charac- 

 teristics of the promyelocyte are still present but 



154 



