In his first publication Ehrlich described the =granules as products of the metabolism of the cells=, deposited within the protoplasm in a solid form, in part to serve as reserve material, in part to be cast off from the cell. On the ground of observations on the liver cells, described in detail in a paper of Frerichs (1883, page 43), Ehrlich gave up this position, though only temporarily. Ehrlich shewed that the liver cells of a rabbit's liver, rich in glycogen, appear in dry preparations as bulky polygonous elements, of a uniform homogeneous brown colour, surrounded by a thin, well-defined yellow membrane. In cells that were not too rich in glycogen, small roundish bodies, clearly of a protoplasmic nature, of a pure yellow, can be seen embedded in the homogeneous cells that are coloured brown with glycogen. "The hyaline cellular ground substance, carrying the glycogen, could not under any circumstances be stained, but the cell-granules above mentioned stained easily with all kinds of dyes. It was further possible to shew by staining that the membrane was chemically different from the granules, since with eosin-aurantia-indulin-glycerine, the membrane stained black, but the granules orange-red." To these observations Ehrlich added the following conclusion, "that the cells of the liver after food really possess a thin =protoplasmic= membrane, and a homogeneous glycogen-bearing substance, in which the nucleus and =round granules= (? functionally active) of protoplasm are embedded. "On comparing these results with those of more recent investigation of the cells, it is easy to determine the location of the glycogen very accurately. Kupffer has shewn, first for the liver cells--and this is now recognised as generally valid--that their contents do not represent a microscopically single substance. In the 'survival' preparation he found, in addition to the nucleus, two clearly distinct substances: a hyaline ground substance in preponderating amount, and a more scanty, finely granular, fibrillary substance embedded in it. Kupffer calls the first paraplasm, the latter protoplasm. On warming the preparation to about 22 deg. C. manifest though feeble movements appeared in the network. It can hardly be doubted, that of these two substances the granular reticulated one--the protoplasm--is the more important; and it should not be erroneous to suppose that the granulations of the network form the =centre of the particular (specific) cell function=. In any case, it is desirable to give a special name, such as microsomes (Hanstein) to these forms, which in the liver cells are recognisable as distinct, round or oval granules, colouring yellow with iodine, and easily and deeply staining in other ways."