Nearly all more recent papers, on the other hand (cp. Loewit's compilation), recognise the =close connection of the blood platelets with coagulation=, first observed by Bizzozero. Whether the substance of the platelets directly yields the material for fibrin formation, as Bizzozero holds, or whether according to the observations on thrombus production of Eberth and Schimmelbusch they play but a subordinate part, is not yet decided. To enter here into the chemical side of this complicated problem, would lead us much too far, and we will only refer to a few clinical observations which illustrate the relations between the clotting power of the blood and the number of platelets it contains. Marked =increase of the blood platelets= occurs in chlorosis (Muir) and in posthaemorrhagic anaemia (Hayem). In both conditions there is a decided =increase in the clotting power of the blood=. In contrast, is the important observation of Denys, who found in two cases of purpura, where as is well-known the =clotting power of the blood is always much lowered= or may even be entirely destroyed, only one morphological blood change, a very marked =diminution of the blood platelets=. Ehrlich likewise had occasion to examine a similar case, in which the blood platelets were entirely absent. * * * * * H. F. Mueller has described a fourth formed constituent of the blood, and given it the name of "=haemoconiae=" or "=blood atoms=," "blood dust." It is found in the plasma of the blood as very small granule- or coccae-like colourless corpuscles, highly refractile, with a very active molecular movement, which keep their shape under observation for a very long time without any special precautions. According to Mueller these bodies are not blackened by osmic acid, and probably contain no fat; they seem to have no connection with fibrin formation, as they always lie outside the fibrin network. Mueller found them in every normal blood, in varying numbers however; much increased in a case of Morbus Addisonii; diminished in hunger and cachexias. More detailed observations are necessary to determine the chemical nature of these forms. Experiments in this direction by extraction with ether, or by the use of fat staining substances, alkanna, Soudan dye, and comparative investigations on lipaemic blood should be successful. FOOTNOTES: [36] The physiological figures found by Brodie and Russell with the aid of this method exceed considerably those of earlier authors. They found a proportion of platelets to erythrocytes of 1:85 or an absolute number of about 635,000 per mm.{^3} LITERATURE[37]. ~Altmann.~ Ueber die Elementarorganismen und ihre Beziehungen zu den Zellen. _Leipzig_, 1 Aufl. 1890. 2 Aufl. 1894. ~Arnold.~ Zur Morphologie und Biologie des Knochenmarks. _Virchow's Archiv_, Bd. 140. ---- Ueber die Herkunft der Blutplaettchen. _Centralbl. f. allg. Pathologie und path. Anat._ Bd. 8, 1897.