"The investigations were performed after Koch's method in the following manner. The fluid (blood) or the parenchyma of the organs (bone-marrow, spleen, etc.) was spread on cover-slips in as thin a layer as possible, dried at room temperature, and after a convenient length of time stained. I had chosen this apparently coarse method for the special reason that for the histological recognition of new, possibly definite chemical combinations, corresponding to the granulations, all substances must be avoided that might act as solvents, _e.g._ water or alcohol, or as oxydising agents, such as osmic acid. In this instance only such procedures may be employed as will leave the simple drying of each single chemical substance as much as possible unchanged." A more detailed study of the process of staining, and of the relation between chemical constitution and staining power, enabled a further advance to be made. And the first result in this previously unworked direction, was the sharp distinction between acid, basic, and neutral dyes, and between the corresponding, oxy-, baso-, and neutrophil granulations. The =triacid solution= was only found after trial of many hundred combinations; and up to the present day this stain in its original form or in slight modifications has played a prominent part in various provinces of histology. The classification of the cell-granules of the blood according to their various chemical affinities which was drawn up by this method is accepted to-day as the most valuable, and the only practical means of grouping the leucocytes. From the first Ehrlich has insisted, =that different kinds of cells possess different granules=, distinguished not only by their tinctorial properties, but also by their various reactions to solvents. It is in this connection indeed, that Altmann's method, consisting of a complicated hardening process, and the use of a single, always similar stain, constitutes a retrograde step, in as much as it tends to obscure the principle of the specificity of each kind of granulation. A further disadvantage of Altmann's hardening method lies in the circumstance, that the cell proteids are precipitated by it in a spherical form, and stain in the subsequent treatment. Hence it is extremely difficult to distinguish what is preformed, and what is artefact. Since A. Fischer's publication, where the formation of granule-like precipitates under the influence of various reagents is experimentally demonstrated, grave doubts as to the reality of Altmann's forms have been raised from various quarters. Ehrlich's dry process, on the contrary, is entirely free from error. Granules cannot be artificially produced, by desiccation, and the stained appearances correspond precisely to what is seen in fresh living blood. The greatest value of the dry method is that the =chemical nature= of the single granules remains unchanged, so that attempts at differentiation are made on a nearly unaltered object[22].