THE PIGMENT MELANIN OF THE SKIN AND HAIR 39 sections. It is perhaps a cross-sec- tion of a dendrite which is running at right angles to the section. According to Masson (28), each dendrite ends in the supranuclear region of a malpighian cell. The dendrites never dip below the epi- dermiscuds line. Sections in which they seem to do so can sometimes be seen,' but serial sections will show that these are artefacts. The proc- esses may spring from any point on the surface of the cells and may be three or more in number and show repeated branching. Often the longest branches spring from the poles of the cell at its longest di- ameter. Since the dendrites are not all in the same plane and often change their direction, it is unusual to get all the processes in the same section or to be able to follow the dendrites to their final termination. The dendrites often seem to com- municate with one another, espe- cially in the dopa sections and some- times form a swelling at their point of junction. The nucleus is large, round and/or oval, contains several nucleoli and is often eccentric. Mitotic figures have not been seen in dendritic cells (Bloch (7), Mieschre (29) and Becker (26), except by Masson (28)). The cell lies between the basal cells, usually at the basal cell level, but sometimes it is seen at a slightly higher level, or almost half of the cell body .may be below the basal line in the cutis. The whole cell, however, has never been seen below the epidermis. The dendrites are ordinarily made visible by their impregnation with melanin, although Miescher has seen cells with dendrites in mela- nomas, when they did not contain any pigment (29). It is a curious fact t.hat the cytoplasm of the den- dritic processes does not take the ordinary stains. This is true of the dendritic cytoplasm in lower animals also (Biedermann (130), Ballowitz (al)). In sections treated with silver nitrate the dendrites are still better seen because their pigment content is brought more to the fore by the reduction of the •ilver (form- ation of silver melanin) which makes visible the fine pale pig- ment granules and perhaps propig- ments. By far the clearest and truest pic- ture is obtained with Bloch's dopa method which demonstrates the pigment-forming oxydase. With the dopa reaction the dendrites are seen either through the content of granular dopa melanin or, better still and as is usually the ease, by the diffuse protoplasmic reaction. This is the only method that clearly demonstrates the cytoplasm, i.e., protoplasm-containing oxidase. In the native as well as in the silver sections the dendrites are demon- strated through their content of pigment granules, which give a pic- ture that is often only outlined by scattered dots, so that it is difficult to trace component parts of the cells (132). With the dopa reaction, how- ever, especially in the diffuse re- action, one obtains a more solid and complete picture. Further- more, in a strong dopa reaction

40 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS nearly every dopa-positive cell may show dendrites, whiie in the native and silver section very few may be seen. The striking and often bizarre forms of fully developed dendritic cells, as they are seen between the other pigmented and nonpigmented epidermal cells, may lead them to appear different from the other epi- dermal cells. They remind one very much of the melanophores of coldblooded animals. Especially in dopa preparations 'they have a striking resemblance to certain cells of the central nervous system. The peculiar morphologic characteristics of the dendritic cells have led many investigators to occupy 'themselves with the problem of their origin and function. In the course of time, various theories have arisen about their genesis and function. It is not necessary in the present state of the problem to discuss the views of the older authors (Schwalbe (7), Rabl (7) and Adachi (25)) who maintained that they are not cells at all but artefacts due to the pig- ment in the intercellular spaces. That the view was false was clearly indicated by the regularity of the demonstration of the nuclei of the cells. There are three main theories which still remain for discussion. Most of the investigators who have occupied themselves with this prob- lem support the view that the den- dritic cells are closely connected with melanin production. This point of view is supported chiefly by the fact that dendritic cells usually contain pigment and are especially numerous during active pigment production. Among the supporters of the theory of the pigment function of the dendritic cells there are di- vergences of opinion about their genesis. Especially the older authors, in analogy to the melano- phores of coldblooded animals, con- sidered the dendritic cells as pig- ment cells of mesodermal origin which had wandered up into the epi•termis in embyonal life, and had remained there to function a melanoblasts (Ranvier (7), Kol- liker (7), Ehrmann (2) and Del Rio Hortega aa)). Still others (Bloch (7), Miescher (29), Kreibich (34)) considered them as cells of ecto- dermal nature which are nothing more than special functional phases of ordinary pigment-building basal cells. According to S. W. Becker the melanoblasts of the epidermis do not rise from the ordinary basal cells (2.6). The French authors (Masson (28), Pautrier, Levy and Diss (35) and ,Caudiere (86)) went still further. They separated the dendritic cells completely from the rest of the epi- dermal cells. The older theory of Masson (19), which was accepted by Pautrier, Levy and Diss, ascribed a metabolic function to the den- dritic cells. As "cellules embo- ceptrices" they have the property of takif•g propig•.nents and various other substances from the reticulo- endothelial cells of the curls into the epidermis and of carrying away again from the epidermis into the