L I E) R. A R. Y 
 
 OF THE 
 
 U N 1VER.SITY 
 
 or ILLINOIS 
 
 610 . 
 V.6 _ 
 
 no \'^ 
 cop. 2 
 
 KLMUIL JdIONAGE 
 
 J 
 
Digitized by the Internet Archive 
 
 in 2011 with funding from 
 
 University of Illinois Urbana-Champaign 
 
 http://www.archive.org/details/melanoblastsmela63zimm 
 
UMIVtv 
 
 The Isocortex of Man 
 
 By PERCIVAL BAILEY and GERHARDT VON BON IN 
 
 Urbana, 1951 
 UNIVERSITY OF ILLINOIS PRESS 
 
610. 
 
 V.6 
 no. 3 
 
 Melanoblasts and Melanocytes 
 in Fetal Negro Skin 
 
 ARNOLD A. ZIMMERMANN, Dr. es Sc, and SAMUEL W. BECKER, JR., M.D. 
 
 Departments of Anatomy and Dermatology 
 College of Medicine, University of Illinois 
 
 ILLINOIS MONOGRAPHS IN MEDICAL SCIENCES, Vol. VI, No. 3 
 
 ^'^^ 
 
 UNIVERSITY OF ILLINOIS PRESS 
 URBANA, 1959 
 
 \m 2o m^. 
 
 u 
 
 UQtCHai. 
 
ILLINOIS MONOGRAPHS IN THE MEDICAL SCIENCES 
 
 is a general title used to comprehend a series of contributions from 
 the fundamental and clinical departments of the Chicago colleges 
 of the University — College of Medicine, College of Dentistrj-, and 
 College of Phannacy. Each issue is an independent publication. 
 For the convenience of bibliographers and librarians, each year's 
 output is called a volume. Each volume consists of approximately 
 450 pages, priced at four dollars (by subscription) for paperbound 
 copies. A volume may consist of one, two, three, or four individual 
 monographs. Prices of individual numbers are indicated in the 
 list of titles given on the back cover. Requests for exchanges should 
 be addressed to the Exchange Department, University of Illinois 
 Library, Urbana, Illinois. All communications concerning sale or 
 subscription should be addressed to the Director, University of 
 Illinois Press, Urbana, Illinois. 
 
 Volumes I, II, and III of this series were issued under the title 
 Illinois Medical and Dental Monographs. 
 
Melcnwhlasts and Mclauocytcs 
 111 Fetal Negro Skni 
 
© 1959 BY THE BOAKI) OK TRUSTKKS OK THK I NIVKHSITV OK ILLINOIS. 
 
 MANUFACTURED IX THE UNiTEi) STATES OK AMERICA. Board of Editors: 
 
 ERNST R. KIRCH, .lOIlN I'. MARBARCIER, SA.MUEL R. M. REYNOLDS, ISAAC 
 SCHOUR, AND RICHARD .1. WINZLER. 
 
 TMK l.mUMd 111' lONCHKSS IIA.-. CAT M.DI^Kll rH\S I'l Kl.UATlO.N AS FOLLOWS: 
 ZlM MLIIMAN N. .\nNOLI) .\lHKI1T. 1895- 
 
 Mil:iiiol)l:isL< :iiul iii('l:iiu)cyt('s in fetal Nf'(!''<) skin |l>yl .Xinold .\. Zilii- 
 liiiiliianii Miul Saiiiiirl W. Hi'ckcr. .Ir. I'rballa, rni\ iM-sily of lllinoi.* Frpss, 
 1959. 
 
 ,59 p. illu>. '.'7 iiii ( llliii(ii> MioM(ii;iai'lis In nualii-al .Ticnrcs, v. 6, no. 3) 
 
 1. ( 'liriiiiiaio|ilioii s. 2. I'jiilirviilony. Unman. 3. Ni'Kio racr. L Hockcr, 
 
 .S.'iiiiiii'l William. 1921 iiiilil aiitlior. 
 
 {^M1,SI.Z5 tUl.77 58-63808 t 
 I.iliraiy iif ( 'niinri'ss 
 
1- '- o 
 
 Contents 
 
 I. Introiluc'tii)n 1 
 
 II. Litoraturo 4 
 
 III. Materials ami Methods 13 
 
 IV. Mi'laiKihlasts and Molaiuicytes in tlic Fetal DiTinis 16 
 
 V. The Development of Distribution Patterns of Ejiidennal 
 ^lelanoeytcs During the Fetal Perioil 20 
 
 VI. Regional Differences in the Frequency Distribution of 
 
 Epidermal Melanocytes in the Negro Fetus 26 
 
 ^'II. Individual Variability in the Population Densities of 
 
 Melanocytes During the Fetal Period 29 
 
 VIII. Discussion 30 
 
 IX. Summary 34 
 
 X. Literature Cited 36 
 
 Plates 39 
 
I. Introduction 
 
 During the past decade, important changes have occurred in the basic 
 concepts of the origin and functional rehitionships of mannnalian and 
 human cutaneous pigment cells. Twenty to 30 years ago, the dominating 
 view was based mainly on Bloch's theory that ordinary basal cells of the 
 epidermis were the pigment producers under appropriate stimuli, and that 
 any true pigment cells encountered in the dermis could have attained their 
 location onlj' by a descent or Abtropjung either in early or later develop- 
 mental periods. 
 
 Pigment formation in man and other mammals thus was placed in a 
 separate category. The essential biological events were said to be different 
 from those known to occur in other vertebrates. That theory failed to 
 account for pigmentary conditions in certain primates and rodents in 
 which the dermis is heavily pigmented and the overlying epidermis re- 
 mains practically free of pigment cells. Adherents to Bloch's theses were 
 inclined or compelled to consider certain normally occurring pigmentary 
 features, such as the so-called Mongolian spots, either as histological 
 curiosities or as having developed from the embryonic epiblast (somatic 
 ectoderm) at a very early period. Xo evidence had ever been given for the 
 latter supposition. 
 
 The newer and most widely accepted views deny that the skin- 
 ectoderm has any pigment-forming potentiality. It has long been known 
 that in other vertebrates, especially amphibians and birds, the source ma- 
 terial for the cutaneous pigment cells resides in the neural crests. These 
 structures constitute paired strands of cells that develop as neurectodermal 
 derivatives on the dorsal sides of the closing neural tube, very early in 
 embryonic life. They are of an evanescent, transitory nature, never exist- 
 ing through the full length of an embryo, but differentiating into several 
 types of cells of which the future cutaneous pigment cell is only one. The 
 potentialities of the neural crest are indeed impressive. 
 
 The credit for experimentally proving that in mammals, too, the pig- 
 ment cells are derived from the neural crest belongs to Rawles ( 1947, 
 1948). Her ingenious grafting experiments showed conclusively that in 
 order to produce pigment, the mammalian epidermis or its hair follicles 
 are wholly dependent on a migratory cell that enters the somatic ectoderm 
 secondarily. Rawles never actually identified the migratory forms of the 
 later pigment cells. 
 
 Zimmermann and Cornbleet (1948) observed the first potential pig- 
 
 This investigation was supported by a research grant (RG 4436, RG 4436C) 
 from the Division of Research Grants, United States Public Health Service. 
 
2 ZIM.MKKMANN AM) UKCKKll 
 
 meiit cells of tlic tVtal Xcfiio opidermis early in the third luoiith. Thej' are 
 not luorlified basal cells, wliicli hefiin to reveal melanin granules late in the 
 fit'tli or in the sixth months — lonj^ after the true pigment producers or 
 melanocytes begin to elaborate melanin. But no one had yet shown that 
 the prenatal human dermis contains any migratory cell stages of future 
 melanocytes. 
 
 If the derivation of iiigment cells from the neural crest was to hold 
 true also for man it obviously became a primary, logical, and challenging 
 task to attempt the identification of their precursor stages in the dermis. 
 This report deals in part witli the differentiation of melanoblasts in the 
 dermis, beginning at 10 and 11 fetal weeks, to melanocytes in the dermis 
 and epidermis. On the basis of extensive material, consisting exclusively 
 of Xegro fetuses, we were able to study a pigmentary cell type that hith- 
 erto had not been, demonstrateil to exist throughout the dermis in the 
 human prenatal period. 
 
 Tlie develoiinient of distribution jiatterns and population densities of 
 pigment-producing cells in the fetal epidermis was also followed in greater 
 detail and with more reliable methods than had been applied before. Cer- 
 tain corrections of published data were made. A strictly statistical treat- 
 ment of the numerical data, however, was not intended. Instead, we 
 deliberately placed the emphasis on new and detailed evidence for the 
 concept that iMgment cells invade the epidermis secondarily. We therefore 
 established their time of arrival at ultimate destinations in various cuta- 
 neous regions. This naturally led to cell counts. In the early fetal periotl. 
 the migration and arrival of melanocj'tes at the dermo-epidermal junction 
 was found to be iiregular. The variability in cell counts of melanoblasts 
 and innnature melaiK)cytes in the tlermis of the third anil fourth fetal 
 months reflects their migration in scattered groups or "swarms." This, in 
 turn, affects the earliest distribution patterns and population densities of 
 melanocytes in the epidermis. A reliable biometrical evaluation of such 
 (lata appeared to be imjiossiljle. The pertinent conditions in the most 
 interesting developmental peiiod were found to be in a highly fluid state 
 (see legend to Talkie 1 ). 
 
 Since a standardized nomenclature for cellular elements concerned 
 with pigmentation is not yet established, we define our terms: A "mel- 
 aniiblast" is strictly an einbiyonic type of cell, potentially capable of 
 proijucing melanin i)ut not containing the fully elal)orate(l pigment. 
 Melanoblasts exist only in the dermis and are first detectable early in the 
 third fetal month. Their round or ovoid cell bodies contain granules that 
 have a specific affinity for reduced silver but cannot be impregnated with 
 ordinary silver nitrate. This is generally accepted to mean that such cells 
 contain a form of "premelanin." Melanin is capable of reducing silver 
 nitrate directly. 
 
MELAXOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 3 
 
 The melaiioblasts, recognized by us from the tenth to twelfth weeks of 
 human development, should be considered as the end stages of a chain of 
 precursor cells that are derived from the neural crest. The earliest migra- 
 tory forms in that chain are undistinguishable from ordinary mesenchymal 
 cells. Since the neural crest of man develops in a cephalo-caudal direction 
 during the second embryonic month, a gap of several weeks remains in the 
 demonstration of the ultimate origin of pigment-producing cells. That 
 gap has been filled by experimentation with mammals (Rawles). There 
 is no reason to doubt that the events are the same in man as they were 
 proved to be in mouse embryos. Certainly there is no evidence to the 
 contrary. 
 
 Melanoblasts rapidly develop one to three short, stubby extensions, 
 the precursors of the future denchitic processes. Thereby they begin to 
 differentiate into melanocytes. Various transitional forms of fusiform and 
 young stellate cells, containing increasing amounts of premelanin granules, 
 make it difficult sharply to distinguish between melanoblasts and melano- 
 cytes. Our definition of the former is definitely more restrictive than that 
 implied by the original term "melanoblasf as introduced by Ehrmann 
 (1885). 
 
 Whether melanocytes begin to elaborate melanin while still in the 
 dermis cannot be used as a criterion for a valid definition. In white skin. 
 for example, the melanin-producing tyrosinase system of melanocytes is 
 generally or temporarily inhibited, except in the sacral region (Mongolian 
 spots). In conditions of vitiligo, the suppression of pigmentary activity is 
 even more pronounced and it is well known that in albinism there is a 
 total lack of the enzyme, althougli the melanocytes are present. Billing- 
 ham (1948, 1949) and Billingham and Medawar (1953) designated such 
 inhibited cells of white skin as "white" or "non-pigmentary" melanocytes. 
 Obviously, a broad definition of melanocytes must rely on purely morpho- 
 logical features. Whether immature or fully differentiated, they usually 
 have fusiform or stellate cell bodies and several processes. Their potential 
 melanin-producing faculty may or may not become apparent. 
 
 All melanocytes are temporarily in the dermis. Some remain there 
 until birth, or, as in the sacral region, until later childhood. The great 
 majority, however, become epidermal melanocytes by an active invasion 
 of the epidermis. In the Xegro. this process begins in the eleventh and 
 twelfth weeks of intrauterine development. While still in the dermis, the 
 fusiform or stellate cells may be designated as immature "dermal" melano- 
 cytes. This working term, of course, is not meant to imply in any way 
 that such cells are of dermal (mesodermal) origin, any more than the 
 epidermal melanocytes would represent a modified epidermal cell. The 
 adjectives "dermal" and "epidermal" are merely used for the sake of 
 
4 ZIMMKHMANN AM) BEIKKH 
 
 brevity ami imlicatc U)catioii as one of the major characteristics of the 
 cells concerned. 
 
 II. Literature 
 
 Tlic vast literature on prohlcins and concepts of cutaneous pipmenta- 
 tioii ill man contains relatively few contributions that deal specifically 
 witii prenatal conditions. We limit our review to studies that concern 
 (a) tlie liistolojiical iilentification of pijjment cells in the pre- and post- 
 natal dermis of man; (b) similar comparative data in other mannnals. 
 particularly in primates; (c) the recognition of prenatal melanocytes in 
 the human epidermis; dli data nii population densities and gradients of 
 distril)iiti()n of such cells; and (e) the experimental evidence for the origin 
 of manunalian melanocytes fiom the neural crest. 
 
 (a) Piyiiieiit cells in the pre- and postnatal dermis of man. The 
 presence of such cells was first established by Baelz (1885), who identified 
 them as early as the fifth fetal month in the sacro-coccygeal region of a 
 Japanese fetus. Baelz consideied the macroscopically visible pigmentation 
 of that area as an important characteristic of the Mongolian race. Hence- 
 forth these pigmented, bluish areas became known as the so-called Mon- 
 golian spots. (Irossly. they had been noticed by Japanese medico-popular 
 writers for some hundred years liefore Baelz's study, and many supersti- 
 tious notions liad been attached to them. Baelz noted also that similar 
 pigmented cells had characteristic relationships to hair follicles, in which 
 they formed a "regular network" innnediately above the pajnlla. 
 
 The first mention in modern science of blue sacral skin areas was by 
 Eschricht (1849 1, who had obtained second-hand information of their 
 occurrence in a full-term Eskimo fetus. 
 
 An important study of pigment cells in the normal corium of Jai)anese 
 fetuses, newborns, and children was made by tirimm (1805). Grossly, 
 he found the blue spots in the sacral region of all newborn Japanese and 
 noticed that the pigmentation in the spots increased during the first post- 
 natal months and began to faile in the second .year. In the dermis of the 
 sacral region Grimm detected the earliest pigment cells in the third and 
 fourth fetal months. In the Mongolian spot of newborns they were large, 
 plump, or elongated fusiform cells, sometimes provided with several proc- 
 esses. The ])rocesses often connected with tho.se of similar iieighi)t)ring 
 cells, and wiiolc "trains" were tiius interconnected, (iiinuu emphasized 
 the irregular tlistrilnition of these cells, often in dense "swarms," located 
 in the deeper two-thirds of the dermis, never in the pajiillary zone. Their 
 length varied between "JO and 50 /i. 
 
 Earliest cell forms of llie tliinl and fonrtii fetal months were round or 
 
MELANOBLAST.S AXD MKLAXOCVTES IX FETAL XEGRO SKIX O 
 
 oval, coarsely granular. Some measured 13 ," in the long axis and 7 /t in 
 width. Grimm conjectured that they originate in very young embryos. 
 He found no evidence for their derivation from connective tissue cells, as 
 had been postulated, and cautiously left the question of their origin open 
 for further embryological and comparative studies. 
 
 Grimm found that dermal pigment cells did not invade any eiMilermal 
 derivatives (sudoriferous and sebaceous glands, root sheaths of hair fol- 
 licles). Although he confirmed Baelz's observation on the arrangement 
 of pigment cells in hair papillae, he could not prove that they were 
 ilirectly derived from the dermis. Grimm's study contains no illustrations 
 of microscopic sections but shows the gross appearance of the blue spots 
 in Japanese children, 
 
 Adachi i lilU3) was the first to identify pigment cells in the dermis of 
 the sacral region in infants and children of the white race. He thereby 
 disproved the opinion that blue spots in the sacral region are characteristic 
 of a specific (Asiatic) race. The development of pigment in the corium of 
 localized ( sacral) areas was recognized as a normal occurrence during the 
 later developmental stages of man in general. 
 
 Adachi studied sacral skin specimens of 76 white individuals, ranging 
 from the fetal period to old age. He made freehand sections of alcohol- 
 fixed material and obtained satisfactory results with unstained prepara- 
 tions. He could not detect any blue spots in fetuses of Europeans and was 
 unable to identifj- dermal pigment cells in white fetuses or in white new- 
 borns. The earliest spots were grossly recognized in one- to three-day-old 
 white infants and dermal pigment cells were numerous in the sacral re- 
 gions of a six-month-old infant and of a 20-month-old child. Adachi de- 
 scribed the pigmented cells in the deeper dermis as fusiform or stellate, 
 measuring 40-80 /<. in length and 4^10 m in thickness. Some cells attained a 
 length of 130 m- He considered them as connective tissue elements and 
 believed that the pigmented cells of the dermis could not reach or pene- 
 trate the epidermis. He was puzzled by a constantly pigment-free (papil- 
 lary) zone between the epidermis and the pigmented layer of the dermis. 
 
 Adachi distinguished two types of connective tissue pigment cells: 
 (a) small cells in the upper layer of the dermis, which obviously corre- 
 spond to the macrophages of modern interpretation, and (b) large cells in 
 the deeper layer. The latter unquestionably correspond to our melanocytes 
 of the dermis and represent the crucial elements of his study. 
 
 Besides "dermal " pigment cells, Adachi described epidermal "chroma- 
 tophores" that are similar in most respects to the dendritic cells or epi- 
 dermal melanocytes. He believed that they were artifacts produced by 
 the arrangement of intercellular pigment granules. In holding this view, 
 Adachi evidently was infiuenced by Schwalbe, in whose laboratory he 
 
I) ZIMMCHMANN AND HKCKKH 
 
 worked in Strassl)!!!": and who. with I'nna. Cohn. HaM. and Krouicyer. 
 was one of the protajioiiists of the thi'orv that cpidciinal "chroniatopiiores" 
 arc not cells at all. Hanvicr. Ilhiinann. i{ichl, and others held, instead. 
 that the epidermal pifinient cells ("chroniatopiiores") were fixed connec- 
 tive tissne cells that had penetrated the epidermis and existed in it inde- 
 jiendently. Khnnann hail introdnced the term ■'melanoblast." refuting 
 thai it i(|)i(s(iitcd either a modified leucocyte, connective tissue cell, or an 
 epidermal cell. .Vccordinji to I'".luinann"s orijiinal definition, "melanoblasts" 
 were dcrixed from the early emhryoiiic mesodermal layer. 
 
 .\da(lii refuted the view that llic .Mongolian spot is an atavistic occur- 
 rence. Since it is found in many colored races and since European children. 
 too. have pigment cells in the corium. the potential blue spot or Kinder- 
 fli'ck had ne\'er been lost in tlie exolution of human pigmentary conditions. 
 Bloch (1901 I, ill a highly siicciilat ivc report unsupported by evidence, had 
 maintained that "the spot is a sort of rudimentary organ which gives clues 
 for the skin color of the ancestors of the yellow race and which one may 
 call a stigma of atavism." 
 
 .Vdaclii licM, instead, that liliic spots are of a i-udimeiitary. regressive 
 character and that this type of ])igmentatioii occurs in all races during a 
 certain period of development. His comjiarative anatomical findings on 
 dermal and epidermal ]Mgmentatioii in jirimates are reviewed under a 
 separate heading. 
 
 Kato (I'.K).")) studied .Mongolian s]iots grossly in r)nn .Japanese children 
 from biith to 1.'^ years, lie gathered statistical data on their localization^ 
 form, size, color, and rate of disappearance. The spots were also studied 
 histologically. He recognized the earliest pigment cells in the dermis of an 
 eight-month fetus. In some skin specimens of children he found pigment 
 cells often adjacent to blood vessels. The cells of a dee]i blue .Mongolian 
 spot ill a iiiiie-iiioiitli-iild chilli were rdiiiid, oxal. nr tiisifdrm. and of a 
 brown color. .Many branched cells ap])eai-ed connected with each other, 
 and though the papillary zone was mostly free of jngment cells. Kato 
 obser\'e(l tluMii occasionally at the (lerinii-e]iiileiiiial junction. He made 
 one of the Hist histological studies of a liliie nexus (adult .lapanese) and 
 identified fusiform, branched, or serpentine cells of brown color in the 
 ileeper two-thinls of the cutis vera. He considered the characteristic jiig- 
 meiit cells ill the dermis of the sacral areas of children and of the blue 
 nexus ill ailiills as iileiitical. (Iriiiiiii had expressed a similar oi>inion 
 xxithiMil gixiiig histological ex'idence. 
 
 Pile gradual disappearance of llie .Mongolian sjiols in later chililhood 
 was e\])iaiiied by I lie disiiit egi at loll of pigment cells and the absorjition of 
 |)iginent granules by dermal lymphatics and venules. In the blue nevus 
 of adults the cells xvere said to remain functional and to i(>tain their 
 characteristic form. 
 
MELAXUIiLASTS AM) MKLANorVTKS IN KKTAL NKliHO SKIN 7 
 
 Bloch (1921) also attempted to fiiul the site and time of the earliest 
 pigment formation in white fetuses. His material consisted of 12 speci- 
 mens ranging from the third to the nintli month. Pigmentary elements 
 within the dermis were encountered in the sacral legion of a fetus of about 
 five months. Bloch described them as "peculiar cells which are not re- 
 lated to the normal pigmentation of the epidermis." These dopa-positive 
 melanocytes in the dermis remained "a puzzle as to their origin and sig- 
 nificance." They were the only exception to Bloch's conccjit of the local- 
 ization of pigment production in strictly ei)itlielial cells of ectodermal 
 origin. Since he detected them only in the sacral area (Mongolian si)ot), 
 Bloch designated them as Mongolenzellen. Nonetheless, he was aware 
 that a complete answer could not be expected from his limited material. 
 
 El Bahrawy (1922) studied under Bloch and tabulated the published 
 reports of Mongolian spots according to their geographical and racial dis- 
 tribution. They were seen macroscopically in newborn and children: 100 
 per cent in Japanese and other Mongolians, SO per cent in Negroes anfl 2 
 to 4 per cent in Europeans. In white children without visible sacral spots 
 the pigment cells in the dermis were pale, scarce, and easily missed. He 
 made histological observations on sacral skin of 112 European cadavers 
 (unstainerl. stained sections, and silver impi-egnations accorrling to Biz- 
 zozzero). Pigment cells in the sacral dermis were first identified in a few 
 white fetuses of four to fiv(> nn)nths and in all specimens from five fetal 
 months to nine-year-old children. In specimens from 12 to 82 years of 
 age, dermal pigment cells were seen in only four, none over 21 years old. 
 The irregular distribution of these cells was noted in many preparations. 
 Oval, pear-shaped, fusiform, or irregularly wa-^'y cells measured 5-10 by 
 30-50 M. 
 
 The argentaffin cells of the dermis were shown to be dopa j^ositive and 
 were recognized, therefore, as true pigment producers or "melanoblasts." 
 Their normal occurrence in European children was considered as a histo- 
 logical or racial curiosity. Bahrawy recognized that phylogenetically such 
 cells might represent "a temporary remnant of generalized pigmented 
 areas in animals, where they exist throughout life." 
 
 In a biopsy sj^ecimen from the forearm of a rhesus monkey, Bahrawy 
 found numerous fu.siform pigment cells in the deeper layers of the dermis. 
 These dopa-positive cells were believed to correspond to those of the 
 Mongolian spot in man. The typical cells of a blue nevus and tlic "Mon- 
 golian cells." as well as the dermal pigment cells of certain monkeys, were 
 thought to be of the same nature. This was an important advance in the 
 direction of a broader concept. Bahrawy speculated, however, that the 
 cells in question might be of early ectodermal origin and had migrated into 
 the dermis during embryonic life. This represents an attempt to support 
 his teacher's theory. 
 
 I 
 
8 ZIMMKIIMANN AMI BKCKKH 
 
 Isliikawa (1!)24) searched for the characteristic piKiiicnt cells of the 
 sacral derinis in 32 Japanese fetuses. The earliest ones were found in two 
 
 sppcinions (of a ^roup of ton fotusos) of tlic early part of the third month. 
 Tlicy wi'ic rt)un(l oi' o\;d cells, nica.'^uiinf; lO-lo n. rarely 23 /i. and showinjj 
 lui processes, (leneially lliey were hifihly scattered and contained yellow- 
 ish-brown pigniont fj''i"iilpsi. Similar cells were identified in the deeper 
 layers of the sacral corinm in four of eijiht fetuses of the fourth month. 
 Tlieie were numerous i)ijimeiit {'ells in the .Monjiolian spot areas of fetuses 
 ranginii fidiii the lit'tli to tlie tenth iiidiith. In specimens older than six 
 months they wei'e mostly spindh'-shaju'd and measured 4-10 by 30-80 f^. 
 
 Ito (l!).")^) reported on dermal piunient cells of the Monjiolian spot. 
 tlie l)lue nevus, and the nevus fusco-coeruleus (of Ota). He stated that they 
 were i)roiUK'e(l l)y "mesenchymal melanohlasts.'" Definite affinities with 
 the nervous system ("sugfjestiuii eiido- or iiei'inevnium"' ) were postulated. 
 Tins concept is reminiscent of Weidenreich's obsolete theory. Ito believed 
 that during its gradual disappearance the Mongolian sjiot becomes es- 
 tranged from the nerxous system, whereas the more localized conditions 
 of the blue nevus and nevus of Ota retain their nervous coimections for 
 life. In his studies of Mongolian spots in Japanese fetuses of varying ages. 
 Ito reported results almost identical witli those obtained by Kato and 
 Ishikawa. As late as 1957 (second report on melanin studies) Ito stiU 
 maintained that melanogenous dendritic cells were derived from peripheral 
 nerve fibers. 
 
 Barry (1952) was the first to report dermal pigment cells in other than 
 the sacral region of man. Working in a French laboratory at Hanoi. Indo- 
 china, he studied fetuses of the yellow race and some scalji specimens of 
 newborn and adult individuals. The earliest dermal melanocytes were 
 identified in silver-impregnated head sections of a 5 cm fetus. We calcu- 
 lated its age to be almost three months. The cells containcMl jiremelanin: 
 some had short processes and were located in the reticulopajnllary zone 
 with their long axis lying parallel to the basement membrane. Some ap- 
 peared to lie migrating toward th(> e])idei'inis. witli processes extending to 
 the baseuKMit membrane. Tliere were as yet no demlritic cells in the 
 epidermis. (French authors persistently designate the latter cell type as 
 "Langerhans cells.") Scalp sections of a fetus of 3.5 months (8.5 cm 
 C.R.L.) showed a few dendritic cells in the epidermis and some dermal 
 melanocytes. In silver-stained scalp sections of a fetus of 4.8 months 
 ( 14.5 cm C.H.L.) there were numerous e)M(lermal dendritic cells and vari- 
 ous forms of dernial nielano('\tes. '{"here \\v\v also dendritic cells and 
 dermal melanocytes in the lumbodorsal-gluteal region. The illustrations 
 of the latter cells resemble our own photomicrographs of similar cell stages 
 in younger Negro fetuses. At five months tlie dendritic melanocytes of the 
 
MELAXOBLASTS AND MELANOCYTES IX FETAL XEfiRO SKIX 9 
 
 epidermis were numerous in the lumbar legiou but the basal cells con- 
 tained no pigment granules detectable with silver. There were many 
 branched pigment cells in the dermis of the scalp, often associated with 
 blood vessels. A similar perivascular arrangement was noted in the scalp 
 of a newborn. The basal cells now contained fine pigment granules, seen 
 both in unstained and silver-treated sections. Xo dermal melanocytes 
 were found in the adult scalp, but the basal cells were now crowded with 
 melanin granules capable of reducing silver nitrate. 
 
 Barry concluded that the dermal melanocytes are the forerunners of 
 the dendritic "Langerhans cells." At first they are without processes and 
 contain only premelanin. Later, branched forms appear, some of which he 
 presumed to become the "dendritic melanoblasts" of hair follicles. In all 
 races pigment granules appear in the basal cells of the epidermis long 
 after the "Langerhans cells" have reachefl their destination. Barry noted 
 that the melanocytes disappear in the dermis when the population of 
 dendritic cells within the epidermis attains a certain density. This is the 
 first suggestion of a puzzling "barrier" effect that seems to regulate the 
 number of invading epidermal melanocytes per unit skin area. 
 
 Barry believed that the dermal melanocytes were derived "probably 
 from the ectomesoderm." without specifically mentioning the neural crest 
 and without being aware of Rawles's work. He postulated that the pig- 
 ment cells of the Mongolian spot, blue nevi, hair follicles, and the dendritic 
 cells of the epidermis originate from the same source. He strongly opposed 
 the theory of Bloch and adduced valid evidence for his plea that it should 
 be abandoned. 
 
 (b) Comparative anatomical data on dermal and epidermal pigmenta- 
 tion in primates and other tnammals. Adachi (1903) made important con- 
 tributions to our knowledge of jjigmentary conditions in the skin of apes 
 and monkeys. In primates, cutaneous pigmentation exists both in the 
 epidermis and in the dermis (orang, chimpanzee), sometimes only in the 
 epidermis (gibbon, spider monkey), or only in the dermis (baboon, bar- 
 bary ape). There were variable amounts of melanin in either layer and 
 the type of cutaneous pigmentation was unrelated to the animal's syste- 
 matic position. In some anthropoid apes (orang and chimpanzee) Adachi 
 observed great numbers of pigment cells in the dermis of practically all 
 body regions. They were spindle- or star-shaped cells, measuring 80-150 
 by 5-10 /i. Similar cells were characteristic for the dermis of various 
 species of Macacus (rhesus monkeys), which have a bluish skin and little 
 epidermal pigmentation. In general the extensor side of the extremities 
 was richer in pigmentation, in either epidermis or dermis, than the flexor 
 side. In the chimpanzee the dermal pigment cells in the extremities often 
 were adjacent to blood vessels. 
 
It) ZIMMKUMAN.N AM) Hi:< KKIi 
 
 Since ill adult man the dcjirco of opidcinial pifinipiitaf ion \'ari('s {jroatly 
 wilii ract', ami the coiium is tree of laifio pijinient cells, Adaclii considered 
 "Homo" as an independent piffmeiitary type. 
 
 Miescher (1922) identified similar pi^nK'nt cills in certain dermal 
 areas of the mouse. They were iiuiikmous around the \il)rissae. where 
 they formed a collarlike network at the upper end of the follicles. Blood 
 vessels were also aecompaiiicd by dermal pigment cells, wliich Miescher 
 described as threadlike, elongatetl, sometimes branched or star-shaped. 
 Their dopa reaction was positive but variable. He therefore recognized 
 them as "autochthonous pigment-producers" and likened them to the 
 pigment cells of the chorioid of the eye. Assuming that the pigment cells 
 in the cutis vera of the mouse, apes, and some monkeys, and those of the 
 chorioid and of the Alongolian spot were of mesodermal origin, Miescher 
 acknowledged that the principle of an ectodermal pigment formation could 
 no longer be maintained or, at least, appeared questionable, 
 
 Daniieel and Cleffnianu (1954) showed that various species of rodents 
 liave pigmentary conditions similar to those in apes and monkeys: the 
 dermis, the epidermis, or the hairs may be the only pigmented structures. 
 They studied embryos and skin specimens of newborn mice. rats, and 
 rabliits. The earliest dermal melanoblasts were identified in mouse em- 
 bryos of 14 to 15 days. The cells were relatively scarce and located near 
 the dermo-hypodermal junction. None were directly beneath the epi- 
 dermis. This had been construed to mean that there is no relationship 
 between dermal and epidermal pigmentary elements. Danneel antl C'let^"- 
 maiin believed, however, tliat dermal melanoblasts rapidly ascend and 
 ])cnetrate the epidermis. They observed that in the ear of mice and rats 
 such migrations occur relatively late, one to three days after birth. In the 
 ear the tlermis retains some pigment cells throughout, whereas in other 
 body regions they tlisappear. On the backs of rabbits only hair follicles 
 contain melanocytes. Danneel and Cleffniaiiii never .^aw a ilirect migration 
 of dermal melanolilasts into developing iiair follicles. The follicles receive 
 their pigment cells from the epidermis during development by migration 
 along the outer root sheath. The migratory process of melanocytes stops 
 soon after birth, and tlic increase in iiuinber of cells dccurs liy rejieated cell 
 di\isioii in thi' papillae^. Tliese investigators were aware of Rawles's ex- 
 l)erimeiital evidence and agree(l witli tlie view that all pigment-producing 
 cells of vertebrates are derived froni ihe neural crest. 
 
 Weissenfels (195(5) contributed interesting details on the earliest 
 jihases of melaiiogenesis in embryos of .Japanese "silky" fowl. The epi- 
 dciinis and featiiers of aduil "silkies" are devoid of i)igiiieiit, l)ut the 
 uiMlcrlyiiig tissues coiilain iiunicrous pigiiuMil-pro(hiciiig cells. Without 
 gi\iiig (liicct cxidi'iicc. the aiillior staled tlial in carlx' ciiilirvonic stages 
 
MELAXOBLASTS AXD MELANOCYTES IX FETAL XEGRO SKIX 11 
 
 "spindle-shaped cells migrate from the neural crest into almost all body 
 regions. " In tissue cultures he observed the origin of premelanin granules 
 within specific cytoplasmic centers of melanoblasts. Granules were pro- 
 duced periodically in waves. The centers were not identical with the Golgi 
 apparatus. Preparatoiy to mitosis the melanoblasts became more spherical. 
 Some processes, crowded with granules, remained connected by slender 
 plasma bridges. After the nuclear events of mitosis, one of the daughter 
 nuclei usually migrated into one of the retained cell processes. No flow of 
 pigment granules could be obser\ed from mother to daughter melanoblasts. 
 The latter, instead, soon began to produce their own premelanin granules 
 from newly arising cell centers. They often formed budlike evaginations 
 of the cell contour from which the demhitic processes arose. These events 
 were observed both in vivo and in vitro, with the phase contrast and the 
 electron microscope. 
 
 (c) Data on prenatal melanocytes in the human epidermis. Pigment- 
 producing cells in the epidermis of Negroes before birth were discovered 
 relatively late in the history of pigment research. The prevailing opinion 
 had been that Negroes were born white and that pigment appeared only 
 during the first few postnatal days, especially at the nail folds, areola of 
 the nipples, and the external genitalia. The concept was based in part on 
 the dicta of renowned histological authorities, e.g., Kolliker, Unna, and in 
 part on careless examination of the hyperaemic, pink skin of the newborn. 
 
 Morison (1889) was one of the first to state that Negro children were 
 born with some cutaneous pigment. In skin sections from the arm of an 
 eight-month Negro fetus he observed pigment in the deepest layer of the 
 epidermis. 
 
 Thomson (1891) identified pigment, in unstained sections of the scalp, 
 in a five-month Negro fetus. He also observed the curved character of hair 
 follicles as well as "interlacing pigment cells" in the hair bulbs. 
 
 Grimm (1895) had seen small amounts of pigment in the rete mal- 
 pighii of Japanese newborn. Adachi (1903) corroborated this and further 
 observed that newborn whites often had a lightly pigmented epidermis. 
 His concept of the epidermal "chromatophore" as a non-cellular structure 
 has been discussed above. 
 
 Bloch (1921) obtained the earliest dopa reactions in the skin of a five- 
 month white fetus. No true pigment was detectable in either epidermis or 
 dermis, but certain melanoblasteiiartige Zellen within the basal layer re- 
 vealed a gray-brown hue. In another specimen of the same age, the matrix 
 of hair bulbs contained "a few cells with processes that looked like 'melano- 
 blasts' and gave a weak, positive dopa reaction." Neither the papillae nor 
 the hair shafts contained any pigment as yet. Fully formed melanin was 
 first identified in a few hair Inilbs of the fifth fetal niontli and in the 
 
12 ZIMMKKMANN AND IIKIKKH 
 
 cpidorinis i)roper of sppcimcns of the sixth to seventh iiioiith. The dopa- 
 positive cells within the basal layer were descritjed as typical "inelaiio- 
 blasts" (in the sense of lihrniann's definition): irregular, star-shaped cell 
 ixxlies with l)ranch('(i inoccsscs. Bloch compared them with "ganglion cells 
 of the brain cortex." He stre.-Jsed again that all pigment was produced by 
 cells of ectodermal origin (basal layer of the epidermis and hair matrix). 
 
 Ziinmciiiiaiin timl ( '(iriil)leet (1!)4S) recognized dendritic melanocytes 
 within tiie ("jjidermis of Negroes early in the third fetal month. They ob- 
 taineil jiositive dopa reactions, exclusively in tliese cells, from the fourth 
 fetal month on. The transfer of melanin granules from melanocytes to 
 neighboring epithelial cells was first noticed late in the fifth fetal month. 
 The papillae of lanugo hairs were seen to contain dendritic melanocytes 
 whose ])rocesses extended directly into the ba.«e of the hair shafts. The 
 melanizatioii of hairs occurred independently of the so-called epithelial 
 matrix, which itself became pigmented later on. The ''interlacing pigment 
 cells" noticed in' Tliomson evidently correspond to the dendritic melano- 
 cytes, which lie between ordinary matrix cells of the hair bulb. 
 
 These conditions were further studied by Zinnnermann (1954). who 
 also attempted the first evaluation of the numerical density of melanocytes 
 in the fetal epidermis of Negroes. The intercellular distances of 200 con- 
 secutively encountered melanocytes were measured in serial sections. 
 These distances decreased by api^roxiniatcly one-iialf Ijetween the fourth 
 and the fifth months. It was concluded tiiat the munber of tlendritic cells 
 had doublet! during that interval. 
 
 Becker and Zimmermaim (1955) canied tiuit numerical analysis fur- 
 ther, ("ell counts weie made in spreads of separated epidermis. In the 
 newborn Negri) they found apjiroximately 1.000 doiia-iK)sitive melanocytes 
 per nun'-. This figure compares favorably with similar counts made by 
 Szabo (1954) in adult white skin. Clold chloiide impregnations revealed 
 the earliest dentritic melanocytes in the epitlermis of white fetuses at six 
 luonths. In Negro fetuses of the third month the first mature melanin 
 gianules were identified in melanocytes of ihe eyelids, the exteriud audi- 
 tory meatus, and s])ecific ai'eas of tlie oi'al nuicosa. 
 
 Ilu. Staricco, I'inkus, and Ft)snaugh (1957) made observations on 
 melanocytes of the prepuce of white and Negro infants. Their illustrations 
 of cells in tissue cultures show many types resemtiling those described in 
 the present study. They found thai "the iclal i\-el\- small ]iiginent cells in 
 I he outgrowth of normal skin explant resemble the early melanoblasts of 
 fetal life rejiorted i)y Zinunermann and ('orni)leet. ' Young pigment cells 
 were recognized in the cultures as bipolar or stellate cells. .\s they maturetl 
 they became strongly dopa positive and showed riclier dendritic ramifica- 
 tions. There were no transitional forms ijetwceii onhnarx' ei)ithelial cells 
 antl melanocytes. Kach cell type gave rise to daughter cells of its own kind. 
 
MELAXOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 13 
 
 (d) Evidence of gradients in the devclop)iient of cutaneous piynienta- 
 tion. Although the migration of nielanoblasts from the neural tube to 
 various body regions had been well established for certain vertebrates, 
 there is only fragmentary information on the rate at which they arrive at 
 ultimate destinations. Hopkins-Fox (1041)) first determined such a 
 schedule in embryos of barred Plymouth Rock chicks. The migratory 
 nielanoblasts could not be identified with certainty in histological prepa- 
 rations. The evidence, therefore, was based on the end results of grafting 
 experiments. Of more than 2.000 transplants, about 1,000 were successful. 
 The earliest migration of nielanoblasts from the neural tube occurred at 
 the level of the mesencephalon, in chick embryos of 8 to 10 somites. In 
 embryos with more than 27 somites "the epidermis overlying all somites 
 tended to yield nielanoblasts upon transplantation." Roughly, an antero- 
 posterior sequence in the migration of nielanoblasts was revealed. In the 
 limb-buds there was a proximo-distal gradient as well as a dorso-ventral 
 migration. The migration was not limited to the epidermis, but nielano- 
 blasts also reached visceral structures along blood vessels (coelomic lining, 
 mesorectum, testes). 
 
 (e) Experimental evidence for the origin of mammalian melanocytes 
 from the neural crest. The fundamental woik by Rawles (1947, 1948) was 
 briefly referred to in the introduction. The pigment-forming potency of 
 various body regions of mouse embryos of a black strain was tested by 
 transplanting them into the coeloni of white Leghorn (albino) chick 
 embryos. Only tissue grafts that contained presumptive or definitely 
 identified neural crest were able to differentiate melanocytes. A medio- 
 lateral spread in that pigment-forming capacity occurred, first at cranial 
 and later at caudal levels. By several hundred grafting experiments Rawles 
 proved conclusively that the somatic ectoderm or its hair follicles are 
 incapable of producing their own melanin. In mammals, too, that faculty 
 belongs exclusively to nielanoblasts and melanocytes, which are derived 
 from the neural crest. 
 
 Such experimental proof is not feasible in man. A search for the 
 earliest phases of melanogenesis by histological means, therefore, appears 
 to be the only possible approach. The results of our own endeavors are 
 presented in the following pages. 
 
 III. Materials and Methods 
 
 One hundred and seven Negro fetuses were collected through the 
 courteous co-operation of various hospitals and institutions: the Depart- 
 ment of Obstetrics and Gynecology of the Illinois Research and Educa- 
 tional Hospitals (Dr. W. F. Mengert), the Department of Pathology of 
 
14 /.IMMKU.MANN ANt) Hl;< KKK 
 
 the University of Illinois. College of Medicine (Dr. C A. Krakower); 
 Cook County Hospital. Departnient of I'atludogy. Chicago (Dr. P. H. 
 Szanto): tiie Carnegie Institution of Washington. Department of Embry- 
 ology, Baltimore (Dr. C Corner); Tulanc liiiversity. New Orleans. De- 
 jiartment of .\natomy (Dr. II. Cununins) and tlie Department of Medicine 
 {Dv. \. l)erl)es); Los Angeles County General Hospital. DeiJartment of 
 Surgical Pathology (Dr. W. Hullock): the Chicago Maternity Center (Dr. 
 Beatrice Tucker), and tlic .Molinc City Hospital. Illinois (Dr. X. T. 
 Braatelien). 
 
 We extend our sincere thanks to all those wiio made this material 
 available to us. 
 
 Al)out (K) specimens were particularly well preservetl. All were 
 formalin-fixed. Detailed records were kept concerning the findings on 
 melanocytes in the dermis and e])idermis. Other specimens were dis- 
 carded liecause tliey were too yoiuig. sIiowcmI varicMis flegrees of maceration, 
 or were otherwise inadequate for our study. 
 
 Age determinations were made from careful measurements of crown- 
 rump length and by applying the formulae of Scammon and Calkins 
 (1929): 
 
 C.R. Length = .(i(j C.H. Length (cm)-|-.5 cm 
 and 
 
 . fC.H. cm ,,.-,. I " , -, 
 
 Age= j — ^g -|-L2o| +./4 
 
 These rules give menstrual age in lunar months of 2S days. 
 
 The ages of several yoimg fetuses obtained from the Carnegie Institu- 
 tion of Washington had been established according to the rigorous criteria 
 of that laboratory. Our own calculations coincided with theirs. We were 
 able to obtain only skin specimens from several older fetuses (twenty- 
 sixth to twenty-eighth week) and newborn Negroes from Cook County 
 Hospital. Chicago. We accepted tlie ages given by the hospital's Depart- 
 ment of Pathology. 
 
 Where\('r feasil)le we made skin shavings from 21 selected body 
 regions of each fetus: two from scalji and cheek, seven areas from the 
 trunk, six dorsal and volar areas of the ujiper limb including the palm, and 
 six posterior and anterior regions of the lower limb, including the sole. In 
 fetuses of the thin! and fourtli months, the fidl thickness of the delicate 
 skin could be usc(|. in older t'cluses, two horizontal slices usually were 
 made, one containing tiie e!)idermis with the upper portion of the dermis 
 and another one consisting of the tleeper dermal layer. By stretching the 
 skin areas of the sjiecimens it was jiossible to make the slices by hand. 
 nicii'ly using a la/or blade, .\fter staining, the jireparations were mounted 
 as spreads, alternately willi tlie dermal or tiie epidermal surfaces upper- 
 
MELAXOBLASTS AND MELAXOt'YTES IN FETAL XE(iRO SKIX 15 
 
 most. Surface examinations of such spreads clearly revealed the number 
 and manner of arrangement of melanocytes. Population densities per mm- 
 could be determined more accurately than is possible in sectioned material. 
 
 Our extensive slide collection was prepared by Dr. Hans J. Knoblich. 
 We are glad to acknowledge our indebtedness for his fine technical assist- 
 ance and faithful co-operation. 
 
 StainiiH/ Tfcliiiiquc. The best results were gained by modifying 
 Masson's impregnation method of ammoniacal (reduced) silver nitrate. 
 After thoroughly washing the formalin-fixed skin slices in distilled water, 
 we placed them in a freshly prepared and filtered solution of 10 per cent 
 ammoniated siher nitrate. Masson's original procedure required 6-8 
 hours impregnation time at room temperature. We incubated our prepa- 
 rations at 55' C from 10 to 30 minutes. Frequent checking of the speci- 
 mens (in distilled water, under the microscope) prevented overstaining. 
 Usually a sepia-brown tint of the skin slices indicated that the incubation 
 could be terminated. Even in such non-cleared preparations the epidermal 
 melanocytes could be readily detected by low power examinations. The 
 quality of the fixation appeared to affect the staining time. 
 
 Adequately impregnated skin spreads, usually about 1 cnr in size, 
 were then treated with a (3 per cent solution of sodium hyposulfite. Gold 
 toning was omitted. The preparations were then dehydrated in an alcohol 
 series, placed in xylol, and mounted in "Permount" as spreads. 
 
 The study of vertical skin sections was accessory. Sections were used 
 mainly for the accurate determination of the depth at which melanoblasts 
 and incompletely differentiated melanocytes were found in the dermis. 
 These sections also were impregnated with ammoniated silver nitrate and 
 incubated at 55° C. The stain was no better, but much faster than that 
 produced by the original ]\Iasson technique. After gold shading, the prep- 
 arations were finished routinely. 
 
 Cell Counts. Cell counts were made in over 500 microscopic fields of 
 epidermal anrl dermal melanocytes. All counts were obtained by means 
 of camera lucida projections. The standard fielcl measured (0.33 mm)-^ 
 1/9 mnr. Each observed melanocyte was traced, and after the field was 
 completely surveyed the sketched-in cells were counted. Their number 
 was then multiplied by nine to obtain an estimate of their population 
 density per mm". Obviously, any error in counting was also multiplied by 
 nine. However, checks were niade by tracing and counting the cells of one 
 particular field on ten different plots. At other times the cell counts in a 
 given field were made by dift'erent observers. In either case the errors 
 were negligible, primarily because the silver impregnations were of high 
 quality and the melanocytes easily identified. A comparison of cell counts 
 in any two fields appeared justified and reliable. 
 
16 ZIMMKHMANN" AMI HKCKKH 
 
 The senior autlmr is responsil)k' fur practically all cell counts and for 
 the analj'sis of the data. 
 
 The photoniicrojiraphs wore inarlc by Mr. Lawroncp Toricllo. Illiistra- 
 tioii Studios, I'liiversity of Illinois. His skillful work is greatlj- appreciated. 
 
 IV. Melanoblasts and Melanocytes in the Fetal Dermis 
 
 \\ c si'arclu'il for i)rccursor stages of melanocytes in many body regions 
 of n-l selected specimens. Many of these fetuses were also used for the 
 study of epidermal melanocytes reported on in succeeding sections. 
 
 Questions of nomenclature were discussed in the introduction. We are 
 using the term "melanoblast" for an embryonic type of cell, potentially 
 able to produce melanin. A "dermal" melanocyte is a more highly differ- 
 entiated cell, fusiform or stellate in shape, containing prenielanin or 
 melanin, and is located in the dermis. It is also designated as an immature 
 melanocyte. An "epidermal melanocyte" is the fully differentiated, 
 dendritic type of cell, also elaborating melanin aiul located exclusively in 
 the epidermis. 
 
 "Dermal" melanocytes become visible in uiisfaiued fetal skin spreads 
 as early as the fourth month (figure 7), Their granules are undistinguish- 
 able from melanin and have affinity for reduced silver nitrate (Masson). 
 
 Doi)a reactions were not feasible in our formalin-fixpd material. 
 
 Tenth and Eleventh Weeks of Fetal Dei'elopment. The earliest 
 melanoblasts were identified in skin specimens of ten Xegro fetuses of this 
 developmental period. Their crown-rump length ranged from 3.4 to 4.5 
 cm. Preparations from various body regions were obtained by stripping 
 small pieces of the delicate skin. In eight fetuses of this group the melano- 
 blasts were found only in the scalp, in others also in the nape and in the 
 saci'al r(>gion. 'i'hey wei'e identified as round cells oi from S to 12 /i di- 
 ameter. Their eccentric nucleus usually contained one or two nucleoli. 
 Fine argentaffin granules were disseminated throughout the cytoplasm. 
 Due to the spherical shape of these cells, they appearcil to be more densely 
 arranged at the jierijilieiy. 
 
 The relatively large melanoblasts were widely tlispersed in the con- 
 nective tissue. .\t that stage of development a true dermis cannot be 
 distinguished from tiie hypoderniis. Cell counts were not made because of 
 the scarcity and wide scatter of these elements. The earliest precursors of 
 future ])igment cells were also recognized in strijiped skin pieces simply 
 mounted in water and examined under the microscope. Such cells contain 
 refractile granules which make tiiem readily iilentifial)le. The size of the 
 refractile granules corresjionded to that of the argentaffin granules seen 
 after im]ir(^gnation with reduced silver nitrate. 
 
MELAXOBLASTS AND MELANOCYTES IX FETAL XEGRO SKIX 17 
 
 Changes in form of the round nielanoblasts occur ah-eady during the 
 tenth week of development. Some were ovoid, assuming the shape of fall- 
 ing drops or of lemons with two small projections at each pole. Gradually 
 more and more fusiform cell types appeared, often of 20 n length. The 
 argentaffin granules tended to accumulate in the tips of the cell processes, 
 giving the impression of active "growth points," Some ovoid cell bodies 
 had two processes at one pole, foreshadowing a tripochil arrangement of 
 future dendritic processes. 
 
 Common to all forms were the argentaffin granules of very fine, even 
 size. Their presence and the continuous series of cell shapes fi'om round 
 to stellate forms were the cytomorphogenic features indicating a single 
 lineage. In scalp spreads of the eleventh week some spindle-shaped cells 
 measured between 30 and 45 m as compared with an average diameter of 
 only 15 /i of the overlying epidermal cells. In general, the differentiating 
 melanocytes of the dermis were conspicuously larger than fibroblasts or 
 fibrocytes of their surroundings. 
 
 During the tenth and eleventh weeks of development, melanocytes 
 in the epidermis were rare. Round nielanoblasts and immature melano- 
 cytes in the fetal dermis, therefore, precede the first appearance of epi- 
 dermal melanocytes (in numbers) by about two weeks. 
 
 Twelfth Week. In nine fetuses of this period, nielanoblasts and im- 
 mature melanocytes were identified in many body regions. We consider 
 this as an indication of the rapid migration or arrival of pigmentary 
 precursor cells from their presumptive source in the neural crest. The 
 crown-rump length of the specimens varied from 5.0 to 6.7 cm. Since 
 the skin was still very delicate, full-thickness strips could be used. ]\Ielano- 
 blasts and transitional forms of "dermal" melanocytes were observed in 
 the scalp, cheek, nape, interscapular, and sacral regions, in dorsal areas of 
 the forearm, of hand and foot, anterior aspect of the leg, and even in the 
 palm and sole. 
 
 The first cell counts of dermal nielanoblasts and of incompletely 
 differentiated melanocytes were feasible. The distribution of "dermal" 
 melanocytes, however, was not uniform through a given field. They ap- 
 peared in groups or "streams" of considerable accumulations. Often the 
 long axes of the spindle-shaped cells were parallel to each other, indicating, 
 perhaps, a directional flow through the connective tissue spaces. 
 
 The cell counts given in Table 1 cannot be taken as an absolute meas- 
 ure of population densities. They merely indicate the relative frequencies 
 with which nielanoblasts and "dermal" melanocytes were encountered at 
 this early age. The table contains the calculated number per mnr of both 
 "derniar' and epidermal melanocytes. The great variability in the number 
 of epidermal pigment-producing cells is due to their irregular distribution 
 pattern at this early age. 
 
18 ZIM.MKK.MANN AM) HKIKKU 
 
 Table 1. Cell Counts of Melanohlast.s anu Kauly 
 Epidermal Melanocytes in the Twelfth Week 
 
 l?(uly regions 
 
 Mclanohlasts and "dermal 
 melanocytes per mm- 
 
 Epidormal melanocytes 
 
 per mni- 
 in corresponding areas 
 
 Scalp 
 
 03, 
 
 182, 
 
 198, 
 
 387 
 
 ' — . 
 
 189 
 
 Interscapular 
 
 
 
 198 
 
 
 90, 144, 
 
 180, 315 
 
 Forearm, dorsum 
 
 
 4o, 
 
 1G2 
 
 
 
 171 
 
 Hand, dorsum 
 
 
 144, 
 
 2.J2 
 
 
 90, 
 
 13.j 
 
 Palm 
 
 
 
 198 
 
 
 234, 
 
 279, 351 
 
 Leg, ant. region 
 
 
 
 126 
 
 
 90, 
 
 36 
 
 Foot, dorsum 
 
 
 
 378 
 
 
 
 18 
 
 Sole 
 
 
 
 162 
 
 
 63, 72, 
 
 153, 243 
 
 The wide range of cell counts in a given region is due partly to the migration and 
 arrival of melanocytes in irregularly scattered groups or "swarms." Stabilized and more 
 typical distrihution iiattcnis of mclandcytes become established in the fourth and fifth 
 fetal months. 
 
 Numerous mieronietric measurements were made with high power 
 magnification. The spherical melanubhists measured from S to 10 m. 
 teardrop or lemon-shaped forms from 12 to 16 ti-. Cells with stubby 
 processes varied between 20 and 25 n. Spindle-shaped forms of immature 
 melanocytes measured from 30 to 45 n in length. 
 
 Figures 3 to (i siiow high power pliotomicrogiaphs of a roinid melano- 
 blast and of early tonus of "deiinal" melanocytes (twelfth week). Obvi- 
 ously, sharp focussing of these cells in the dermis of skin spreads is more 
 (Ufficult than in sections. 
 
 Thirteenth and Fourteenth \\'( el.s. The six fetuses of this fetal period 
 varied in crown-rump length from 7.0 to 0.5 cm. Melanoblasts and 
 "dermal" melanocytes now were present also in the pectoral region and on 
 the anterior aspects of arm and thigh. Figure illustrates early forms of 
 melanoc\'tes in tlie dennis of a lumbosacral spread. 
 
 Cell counls varied gencrall.N' helween 200 and 4(11) jier mm-'. In one 
 fetus, however, the counts wcvv nuicli higher in the sacral region and in 
 the dorsum of hand and foot. Specifically, the comits were ()S4 and 1.305 
 for the sacral region, 003 and 1.101 for the dorsmn of the hand, and 855 in 
 the ilnrsuiii (it tlie foot, ('nunts of tiic sanu' cell types in the palm and 
 sole were the lowest for tlmt specimen: lOS and 201, respectively. 
 
 During the thirteenth week yoimg ei)idermal melanocytes appear in 
 many skin regions. Sudi juvenile forms usually are slender, fusiform cells 
 with long |)riniary dendritic processes. They are easily recognized in a 
 liiglici' focal plane lli:in llial of the "dernial" inelan(ie>'tes. The latter also 
 
MELANOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 19 
 
 teiul to be spindle-shaped with cell processes just beginning to form. Fusi- 
 form "dermal" melanocytes from the dorsum of the foot are shown in 
 figure 8. 
 
 Fifteenth and Sixteenth Weeks. The seven fetuses of this develop- 
 mental period, terminating the fourth month of pregnancy, varied in 
 crown-rump length between 9.5 and 11.7 cm. The population density of 
 "dermal" melanocytes, in general, was higher than in the preceding weeks, 
 although there still was considerable variation. This may have been due 
 partly to technical procedures. If surface shavings are too thin the prep- 
 arations may reveal only part of the population of dermal melanocytes. 
 Cutaneous areas from the scalp, and dorsum of the hand and foot, some- 
 times contained 500 to 700 melanoblasts and "dermal" melanocytes per 
 mm-. Spindle-shaped forms usually were predominant, although in one 
 fetus of 16 weeks many round melanoblasts were typical of various regions, 
 particularly in the dorsum of hand and foot. 
 
 During this developmental period, "dermal" as well as epidermal 
 melanocytes become visible in u)}stai})ed preparations. Figure 7 shows 
 spindle-shaped cells in the dermis of an unstained scalp spread (fifteenth 
 week.) The presence of true melanin in those cells was proved by impreg- 
 nations with ordinary silver nitrate (Bizzozzero's method). Relatively 
 young forms of "dermal" melanocytes from the dorsum of the hand are 
 shown in figure 10. 
 
 Seventeentli to Twentieth Weeks. Split-skin preparations were made 
 of 15 fetuses of the fifth month of pregnancy. Their crown-rump length 
 varied between 12.5 and 16.5 cm. 
 
 Cell counts of "dermal" melanocytes remained high in most regions. 
 There were over 1000/mm- of such cells in the sacral region (Mongolian 
 spot) of two specimens (figure 11). Counts of epidermal and "dermal" 
 melanocytes of a preparation usually were lower by one-half or two-thirds 
 for the latter type of cells, except in the sacral region. Cell counts in 
 preparations impregnated with reduced silver were consistently higher 
 than those in unstained skin spreads. This agrees with the generally ac- 
 cepted view that ammoniacal silver nitrate reveals precursor stages as well 
 as fully elaborated melanin granules. In unstained preparations presum- 
 ably only the latter are visible. 
 
 In some fetuses of this period, the spindlc-shapetl melanocytes in the 
 dermis were the most characteristic cell type. They had become longer 
 and measured from 40 to SO /i. In other specimens the majoiity of the 
 "dermal" melanocytes were of the round or ovoid form. They may have 
 been "young arrivals" in the particular regions (lumbosacral, dorsum hand, 
 dorsum foot). 
 
 During this developmental period, immature melanocytes also tended 
 
20 Zl\lMi:iiM.\N.\ AMI HK(Ki;H 
 
 to becuint' adluTcnl Ui small Ijluud vessels. Whole chains clearly outlined 
 the course of caijillaries. The cells often had the ajipeaiance of silver- 
 impregnated "pericytes." Figure 12 ]ioi trays such an arrangement in the 
 dermis of the scalp. Migratory melanocytes may also follow the course of 
 nerves in the dermis. Such migration, however, seems to be incidental 
 rather than essential. Similar observations by Ehrmann may have induced 
 him to bclicNc lliat melanoblasts were derived from the adventitia of blood 
 vessels. 
 
 Sixtli Muidlt to Birth. The mateiial for this period consisted of three 
 specimens of the twenty-sixth and twenty-eighth weeks and of four full- 
 term fetuses, ill llic full-term fetuses the epidermis was sloughed; the 
 dermis, however, was well pieserved and excellent silver impregnations of 
 "dermal" melanocytes were obtained. 
 
 Characteristic of this late fetal j^eriod is the disappearance of the 
 melanoblasts and melanocytes from the dermis of most skin areas. Per- 
 haps the cells merely lose the active enzyme system that is necessary for 
 the elaboration of melanin. They would then become undetectable with 
 (lur present methods. They may disappear entirely, though we cannot 
 be sure. 
 
 E(iually characteristic, however, is the fact that "ilermal" melanocytes 
 remain a constant feature of the scalp, the sacral region, and the dorsum 
 of the hand and foot. In the last two regions their number varied between 
 500 and (iOO cells at six months as well as at birth. In the sacral region of 
 several newborn specimens we counted from 900 to 1,400 "dermal" mela- 
 nocytes per mm.- Figure 13 shows such cells from the dorsum of the hand 
 (if a iiewlioru Negro. 
 
 V. The Development of Distribution Patterns of Epidermal Melanocytes 
 
 During the Fetal Period 
 
 It has been known for some years {Zimmermann and Cornbleet, 194S) 
 that the first epidermal dendritic cells appear early in the third fetal 
 month. Beckei' and Zinuiieiiuaim (10.").")) had also establisluHl that in the 
 iiewliorii Xegro theie are approximate]}' l.UOO melaimcNies per mm'" of 
 epideiinis. But only incomplete information has been available concerning 
 the i)o])ulation density of such cells at different fetal periods and in various 
 skin areas. 
 
 Previous data wnv based on skin sections in which the mean distance 
 between 200 consecutively encountered melanocytes hail been measured. 
 The ])roi)ai)le number of melanocyt(>s jier nun" was then estimated. It is 
 difficult to identify fuiictidiially immature dendritic cells in skin sections 
 of earlv fetal months, in trviiig In avoid llial the same cells were counted 
 
MELAXOBLASTS AXD MELANOCYTES IX FETAL XEGRO SKIX 21 
 
 ill adjacent sections, we had measured only alternate sections. This pre- 
 caution led to cell counts that were too low. Our more reliable, present 
 method calls for correction of the figures published in our 19.55 study. 
 
 Forty-three fetuses of our material had been aborted between the 
 tenth and twentieth weeks of development. This developmental period 
 proved to be an important one for the differentiation of epidermal mela- 
 nocytes and the establishment of their regular distribution pattern. No 
 intraepidermal dendritic cells could be seen before the tenth week (fetuses 
 with a C.R.L. of 40 mm or less). In the twentieth week, however (C.R.L. 
 of over IBO mm) the population density of epidermal melanocytes in most 
 skin areas already resembled the conditions at birth (800-1,100 per mnr). 
 The following is an account of the development of distribution patterns of 
 dendritic cells until birth. 
 
 Tenth Week. The earliest epidermal melanocytes were identified in 
 the interscapular region of a fetus measuring 34 mm in crown-rump length. 
 Its age was calculated at 2.7 lunar months or 10 3/7 weeks. Widely scat- 
 tered dendritic cells were observed, some with fairly long, branched proc- 
 esses. Figure 14 shows one of these earliest intraepidermal cells. Three 
 dendritic processes extend through a total distance equal to about ten 
 ordinary epidermal cells. Argentaffin granules of premelanin became vis- 
 ible with Masson's reduced silver technique. They could not be seen in 
 unstained preparations nor after treatment with ordinary silver nitrate 
 (Bizzozzero). Fully elaborated melanin, therefore, was not present. The 
 dermis contained numerous round melanoblasts, e.g., in the sacral region. 
 Counts of the sporadically appearing, earliest melanocytes of this stage 
 would be meaningless. 
 
 Eleventh Week. Scattered intraepidermal melanocytes were observed 
 in various cutaneous areas of eight fetuses of this period. Their crown- 
 rump length varied between 37 mm and 55 mm. Scalp, cheek, and nape 
 revealed dendritic cells in most of these specimens, although never in high 
 numbers. A few cells were also identified in the interscapular, para- 
 umbilical, and gluteal regions. Even the epidermis of the delicate limbs, 
 on the dorsum of the arm and the posterior region of the thigh, contained 
 a few young dendritic cells. In most areas there were fewer than ten early 
 melanocj'tes per mm'. Cell counts were not reliable. The youngest forms 
 were fusiform with slender, drawn-out cell bodies and two dendritic proc- 
 esses. Some were triangular, with indications of three processes, or even 
 with secondary branches. Some of the processes measured 50 p. in length. 
 The total length of the fusiform cells usually ranged between 50 and 75 n. 
 The distal limb segments were still devoid of dendritic cells. 
 
 TiceJfth Week. Epidermal melanocytes now appeared also in palm 
 and sole. The distribution of early dendritic cells was regular enough to 
 
22 ZIMMKUMANN AM) HKCKKK 
 
 allow coll (tiiiiits ill five of the seven available fetuses of this period. Their 
 ciowii-niiiii) Iciifitlis \aiii'(l Ix'twecii .">() and ()7 imn. Tlip cell coiiiifs ranged 
 from •'iO to L'(M). In two specimens, however, the population density of 
 youufi dendritic cells already was remarkably high, e.g.. 576 cells per mm' 
 in the nape and 738 cells per mnr in the sole of one. (503 cells in scalp and 
 513 in the palm of the other specimen. Since the distribution pattern was 
 uneven, tiiese counts convey at least an impression of the rapid migration 
 of melanocytes into specific epidermal areas. Transitional cell forms be- 
 tween young dendritic cells and ordinary ejiithelial cells were not seen. 
 Each of the 20 .skin areas contained .some dendritic cells at this early age. 
 Their cytodif!'erentiation occurred rapidly. Figures 15 and 16 represent 
 early cell forms and distribution patterns of epidermal melanocytes of a 
 twelfth week fetus. The fusiform cell bodies of the sole, showing long 
 primary dendritic processes, were characteristic of young dendritics in 
 all cutaneous areas. In the sole they had just appeared at this age; in 
 the nape (figure 16) they had already attained a higher degree of 
 differentiation. 
 
 Figures 17 and IS re))resent cell stages encountered in full-tiiickness 
 skin spreads of two fetuses obtained from the Carnegie Institution of 
 Washington. Depaitment of Embryology (Baltimore). Their respective 
 ages hail been determined in that laboratory as 12 weeks (Carnegie speci- 
 men Xo. 9014) and 12 3/7 weeks (Carnegie specimen No. 8613). Figure 
 17 shows highly differentiated melanocytes in the epidermis of the scalp 
 and figure 18 of the anterior region of the leg. The surface views of these 
 spreads show the nuclei of ordinary epidermal cells and permit a com- 
 parison with the length of the dendritic processes. The dendritic cells 
 reveal accumulations of argentaffin granules to the tips of their processes. 
 
 Figures li) and 20 portray conditions of melanocyte differentiation in 
 the najie and in the dorsum of the arm. The cells are of a uniform type 
 in the formei- liut varialile in the latter region. The fusiform, slender cell 
 forms prol)al>!y had recently "aniveii" in the epidermis on the dorsum 
 of the arm. 
 
 A relative "timetable of arrival" of ejiidermal melanocytes in differ- 
 ent skin areas of an individual is siu)wn liy cell counts. During the twelfth 
 week they were consistently higher in the palm than in the .«ole. e.g.. 
 351:72; 234:63 per nun-'. 
 
 Thirteeiilh mid Fourteenth W'icLs. A markeil increase in tlie numl)er 
 of epidermal nielanoc^'tes chaiacterized lliis jiiTind. In the palm api^eared 
 the first indications of rete ridges (ejiidermal crests). 
 
 We studied six fetus(>s ranging from (it) to 75 nun C.H.L. in the 
 thirteenth week and from S5 to 133 nun in the fourteenth week. 
 
 Tiie distrii)Ution of Nounu dendritic cells was more imiform and cell 
 
MELAXOBLASTS AND .MELANOCYTES IN FETAL NEGRO SKIX 23 
 
 counts became reliable. Cell distribution remained irregular in only a few 
 skin areas. The orientation of melanocytes often indicated a directional 
 flow or migration ( figuie 20). Population densities of tlie dendritic cells 
 varied, both regionally and individually, between 300 and 1,000 per mnr. 
 In the scalp, cheek, dorsal trunk regions and dorsum of forearm the counts 
 were between 800 and 1,000 cells. Great waves of rapidly differentiating 
 dendritic cells thus appear to arrive at their destination during a short 
 period of only two or three weeks. Their population density ([uickly ap- 
 proached that of the newborn. Since nuclear events in epidermal mela- 
 nocytes cannot be observed after sih'cr impregnations, we are unable to 
 say whether the increase in cells was due, in part, to mitotic divisions. 
 
 Figure 21 illustrates the distribution pattern in the scalp of a Negro 
 fetus of 131/2 weeks (75 mm C.R.L.). We counted 936 dendritic cells per 
 mm- of that area. 
 
 The distrilnition features in palm and sole were of special interest. 
 When they first appear the melanocytes are fairly evenly scattered 
 throughout the epidermis of palm and sole. Figure 23 shows that pattern 
 in the palm at 13 weeks. It changed rapidly with the differentation of 
 rete ridges (epidermal cristae) between the thirteenth and seventeenth 
 weeks. This occurs first in the palm. Figure 24 shows the dendritic cells 
 located almost exclusively on the rete ridges. Palm and sole contain some 
 melanocytes until birth and even to adult life. Their potentiality to form 
 melanin, however, appears more inhibited than elsewhere in the body. 
 
 Fifteenth and Sixteenth Weeks (end of fourth month). Our material 
 consisted of three fetuses of the fifteenth week (95 to 105 mm C.R.L.) and 
 of four fetuses of the sixteenth week (107 to 117 mm C.R.L.). In general, 
 the distribution patterns of the dendritic cells were regular and uniform 
 in all skin areas, except in the palm. Cell counts were higher than in 
 preceding periods. This reflects a continued infiltration of the epidermis 
 by melanocytes from the dermis. At 15 weeks the population density of 
 dendritic cells varied between 600 and 1,200 per mm-. Figure 22 shows a 
 characteristic distribution pattern of epidermal melanocytes in the inter- 
 scapular region at that time. 
 
 In unstained spreads of scalp specimens of the fifteenth and sixteenth 
 weeks, the dendritic melanocytes were faintly visible. The degree of their 
 differentiation corresponded to that shown by impregnations witli am- 
 moniacal silver nitrate. Unstained dendritic cells were also identified at 
 the palpebral fusion plate. Impregnations by Bizzozzero's method proved 
 that they contained fully elaborated melanin. We had previously shown 
 that melanin is first formed in melanocytes of certain head regions early 
 in the fourth month. 
 
 The rete ridges in the jialmar epidermis become well established during 
 
24 ZIMMKUMANN AM) MIOCKKU 
 
 the fifteenth week. The previously scattered luehiiiocytes now were located 
 on the ridges and their nunil)er decreased to l.")()-20() cells per mm-". This 
 may he due lo a loss of staiiiahilily or to actual disappearance of the 
 dendritic cells. Duiiiiii ihc fit'tccntli week tlie sole usually has no rete 
 ridges as yet. 
 
 During the sixteenth week, cell counts of epidermal melanocytes varied 
 between 700 and 1,200 per nun'. The palmar and plantar areas were again 
 the excei)tions. In the palm all tlie dendritic cells were on the rete ridges 
 and nunil)cr(Hi between loO and 2.")() cells per luni". In the sole, instead, 
 the epidermal cristac were just bciiinning to form and tiie number of 
 the evenly scatter('(| melanocytes remained lelatively high: 37.) to 450 
 per mm-. 
 
 Sev€iit(ciitli and Eiyhtcenth H'crA*-. There were no important changes 
 in this developmental period. Specimens of the seventeenth week meas- 
 ured between 125 and 129 nun in C.R.L. and lietween 130 and 137 mm 
 C.R.L. in the eighteenth week. 
 
 In general, the po]nilation density of melanocytes was similar to that 
 at the end of the fourth month of development (600-1,200 cells per mm"). 
 There were two exceptions. Counts in skin spreads of a 130 mm fetus 
 were consistently low (300-700 cells). There were some signs of relatively 
 poor preservation and the counts may not be reliable. I'nusually high cell 
 counts were obtained in preparations from a fetus of 17 weeks; manj' ex- 
 ceeded 1,000 per mm-'. I'ixation and silver impregnation were good. We 
 ascribe such high counts of dendritic cells to individual variation, which 
 will be discussed in a subsecjuent section. 
 
 During the seventeenth and eighteenth weeks, the development of 
 rete ridges in palm and sole had further ])rogressed and constituted very 
 characteristic features. The melanocytes were locatetl exclusively on the 
 ridges. Theii' number was consistently higher in the .st)le than in the palm. 
 Cell counts in both areas were higher in specimens of the seventeenth than 
 in the eighteenth week. In the palm the average number per mnr de- 
 creased from about 200 to 100. in the sole from al)out 400 to 175 cells. 
 This may l)c due to a loss in functional activity with resulting unstain- 
 altility, or it may mean an actual decrease in the numlier of melanocytes. 
 
 Figure 24 shows the arrangement t)f the melanocytes on the epitiermal 
 ridges in the palm of a Negro fetus of 17 weeks. There were 126 dendritic 
 cells per mm' (compare with figures 23 and 26). 
 
 X'nu'Ueiith and 'rwcnlicth Weeks (end of fifth month). Split-.^kin 
 ])re])arations were ol)tained from five fetuses l)elonging to the nineteenth 
 week of development (145 to l.")!) mm C.R.L.) and of two fetuses of the 
 twentieth week (165 nun C.R.L.). Cell counts fiom one fetus of the 
 twenty-first week were included in this jteriod. 
 
MELAXOBLASTS AXD MELANOCYTES IX FETAL XEGRO .SKIX 25 
 
 The population densities of epidennal melanocytes varied between 500 
 and 1,000 cells per mm'-. In one fetus it ranged from 700 to 1,100 and in 
 another between 800 and 1.400, 
 
 The impression gained from an analysis of cell counts in the seven- 
 teenth and eighteenth weeks was sustained: the great influx of melano- 
 cytes into the epidermis had occurred before the end of the fourth month. 
 There was now a distinct slowing of that process, A degree of stabilization 
 appeared to be attained, although various cutaneous areas still contained 
 melanocytes in the dermis, Barry ( 1953 ) had noticed it but nothing is 
 known of the causes. We can merely state that the period of penetration 
 of the epidermis by melanocytes is relatively short and occurs essentially 
 before the midpoint of pregnancy. 
 
 Figure 25 shows a typical distribution pattern of epidermal melano- 
 cytes in the anterior abdominal wall at 19 weeks. The population density 
 was 873 cells per mm'-. 
 
 The gradual decrease in the number of melanocytes on the epidermal 
 cristae of the sole is shown by the following figures: 531 per nmi" at 17 
 weeks, 360 at 19. and 270 at 20 weeks. 
 
 Twenty-Sixth Week. We were able to obtain skin specimens of various 
 body regions from two fetuses of 6.6 months. The counts of epidermal 
 melanocytes from 10 cutaneous areas of one fetus (220 mm C.R.L.) varied 
 between 700 and 1.100 per mm-. In the sole there were only 81 dendritic 
 cells per mnr, all located on the rete ridges. In the other fetus of this 
 period (225 mm C.R.L. ), the cell counts from seven cutaneous areas ranged 
 between 1.000 and 1,485. 
 
 These counts are in good accord with our findings from the end of the 
 fifth fetal month and with those in the newborn. They indicate that no 
 additional wave of melanocyte "arrivals" occurred during the second half 
 of the fetal period. Scattered melanocytes from the dermis may neverthe- 
 less enter the epidermis in small numbers. 
 
 Full-Term and Negro Injants oj the Xeonatal Period. The epidermis 
 had been sloughed off in skin specimens of four full-term fetuses. In the 
 palm of one specimen we were able to count 288 melanocytes per mm". 
 Figure 26 shows their arrangement on epidermal ridges, between openings 
 of the sweat pores. In the sole of another specimen there were 171 den- 
 dritic cells per mm-. Although unsatisfactory for an over-all study of pig- 
 ment cells in the epidermis, the four fetuses nevertheless revealed great 
 numbers of melanocytes in the dermis of special areas. Reference to those 
 findings was made in another section. 
 
 Reliable counts of melanocytes in the newborn had been previously 
 reported from the abdominal wall (Becker and Zimmermann, 1955). Epi- 
 dermal spreads had been obtained by trypsin digestion. They were fixed 
 
2() ZIMMKHMANX AM) HKCKKH 
 
 ill 2 i)cr cent tnriiialiii tnr t'oiii' Ikiius ami llicii iiiciibatecl in a 1 per cent 
 (lojja solution for three hours. Counts ranged between 083 and 1.152 dopa- 
 positive denthitic cells per nini". In a ten-day-old Xegro infant, the dopa- 
 treated ejiiderinis contained between ()17 and S14 melanocytes per iiinr of 
 
 abiloiiiiiial surface area. 
 
 VI. Regional Differences in the Frequency Distribution of Epidermal 
 Melanocytes in the Negro Fetus 
 
 A total of 412 field counts of ejMdernial melanocytes were made. Only 
 such preparations were used in which the melanocytes were well impreg- 
 iialed witii rerjuced silver nitrate. There were differences in the intensity 
 witii which iii(H\iihial cells took up the silver. Since all preparations were 
 made with ceiual care, our numerical data are considered adeiiuate for an 
 estimate of regional tlifferences in population densities. They represent a 
 first attempt at determining the time of arrival of fetal melanocytes in 
 various body regions. Wherever feasible 20 cutaneous areas of each fetus 
 were studied. 
 
 Figure 1 graphically illustrates tlie pooled data for two developmental 
 periods: (a) the twelfth week, when the epidermis is being invaded by 
 melanocytes in numbers, and (b) a longer interval from the thirteenth to 
 the twenty-first weeks (end of fifth fetal month). During that time occurs 
 a stabilization of the cell counts. ( 'onsec|uently. the population density of 
 epideinial melanocytes leinaiued relatively constant diu'ing the second half 
 of pregnancy. 
 
 During the twelfth week, cell counts in ilifl'erent regions varied from 
 about 20 to 400 dendritic cells per mnr (see Table 1). A total of 50 
 counts were made. Tlie highest counts were obtainetl from the scalp, 
 cheek, nape, and interscapular regions. The counts for various cutaneous 
 aieas of the upper limb were higher than tho.^e of the lower limb. 
 .Surprisingly high numbers of dendritic cells were obtained for palm 
 (351/mnr) and sole (243/nmr). 
 
 Since the head region of an embryo and young fetus grows earlier and 
 faster, and the anterior limb differentiates ahead of the jiosterior limb, the 
 population densities of nielaiiocytes appear to conform with the general 
 se<iuence in cephalo-caudal growth. .\\\ anteio-jiosterior gradient in the 
 distril)ution density of early melanocytes was clearly indicateii. 
 
 Relatively high counts in the jialm and sole may be due to an accu- 
 niiilalioii (if iiiigraloiy niclaiioblasts in distahnost ar(>as. Tlieir in\-asion 
 of tiie ei)ideiniis and differentiation into ileiidritic c(-lls may occur at a 
 high rate. The iiopiilation density in tiie palm and sole increased liuriiig 
 the thirteenth week but decreased tliereafter. .V dorso-ventral difference in 
 the iiiiiiilier o\' iiielaiiocN'les was iml e\id('ii1 in the Iwcltth week. 
 
MELAXOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 27 
 
 COMBINED AVERAGE COUNTS OF DENDRITIC MELANOCYTES PER MM 
 
 HOG 
 1000- 
 900- 
 800- 
 700- 
 600- 
 500- 
 400- 
 300- 
 200 
 
 100 
 
 Heod ond Trunk 
 13th -21 St fetal week 
 
 (TOTAL COUNTS^ 270) 
 Upper Limb 
 
 I2th fetal week 
 
 Lower Limb 
 
 
 
 
 J„ 
 
 
 Figure 1. Population densities of ejiidermal melanocytes in various loody regions. 
 The lower curve shows average counts of melanocytes per mm- during the 
 twelfth fetal week. The upper curve represents pooled frequencies between the 
 thirteenth and twenty-first fetal weeks. Tlie number of melanocytes in palm and 
 sole dro]3s sharply after tlie thirteenth week. 
 
 The top portion of the graphs in figure 1 shows the average frequency 
 distribution of epidermal melanocytes in 20 areas of fetuses between the 
 thirteenth and twenty-first week. It is based on 270 field counts of den- 
 dritic cells. 
 
 The data were first plotted separately for the thirteenth week (25 
 counts), for the fourteenth to sixteenth week (149 counts) and for the 
 nineteenth to twenty-first week (96 counts). Since there was an overlap of 
 population densities in corresponding areas we pooled the entire group of 
 data. The graph portrays the main trend of regional cell frequencies dur- 
 ing the fourth and fifth fetal months. 
 
 In fetuses of the thirteenth week the counts of epidermal melanocytes 
 were significantly lower for the lower limb as compared with those of the 
 upper limb. This agrees with the slower or delayed rate of development 
 of the hind limb. In the following few weeks that difference in population 
 densities became rapidly erased. 
 
 Between the thirteenth and twenty-first weeks the average counts of 
 
28 ZIMMKH.MANN AM) HKIKKK 
 
 luclaiiocytt's ltd' all culaiicuus icfiitjiis xaiied ht'twciMi <).")() anil it.jU iimi". 
 The greatest influx of epklennal inelanoeytes into any of these areas 
 had occurred Ix'twccn the twclt'ih ami fouitccntli weeks of intrauterine 
 (levelopnicnt. 
 
 The most characteristic features were the gradual disapiiearance of a 
 cephalo-cauilal gradient, and the clear emergence of a dorso-ventral gra- 
 dient ill jxipulatidn densities, tor the trunk. ui)i)er and lower limb. 
 
 Till' cdunts (if melanocytes ilid luit icmain significantly higher in the 
 (■(>plialic portions of the fetus. High counts were obtained from the sacral 
 region. The number of dendritic cells in dorsal areas of both limbs was 
 also approaching the previously higher counts in the head region. 
 
 The ])ool('d axcragc counts fiDin the doisal areas of trunk and limbs 
 (littered from those of corresjionding ventral areas. They were significantly 
 higher in the sacral area, dorsum of arm, forearm and hand, posterior 
 region of thigh, and in the calf, than those of the paraumbilical and pec- 
 toral regions, or of anterior areas of arm, forearm, thigh, and leg. This is 
 clear evidence of a dorso-vential giadient in tlie pojudation densities of 
 melanocytes at that fetal period. 
 
 The high counts in the sacral legion (average: !)47/mm-) are of 
 special interest since that area corresponds to the caudal end of the neural 
 tube and of the neural crests. Great numbers of pigmentary precursor 
 cells accumulate in the dermis of that area. They constitute tlie substrate 
 foi' the so-call('(i .Mongolian s])ot. .Many of tiies(> cells may later penetrate 
 the epideiniis and thus lead to a liigh po])ulation density of dendritic cells 
 in the sacral area. All the epidermal melanocytes of the lower limb, inci- 
 dentally, must also be derived from the dermal pool of melanoblasts in 
 that area. 
 
 Sjiecial features prevailed in the epidermis of palm and sole of the 
 fourth and fiftli months. The first epideiinal cristae or rete I'idges appear 
 ill the i)alni about two weeks earlier than in the sole. They were easily 
 identified in both areas during the eighteenth week. None were present 
 elsewhere in the skin of that period. During the formation of rete ridges, 
 the counts of palmai' and jilantar melanocytes dropped sharply. Their dis- 
 li'iliutioii ])att('in changed from one of cNciily scattered cells to one of 
 restricte(| alignments on the rete ridges. There remained only rare den- 
 dritic cells Ijetween the epidermal crests. This decrease is perhaps more 
 simulated than real; if the enzymatic activity necessary for melanin pro- 
 duction lu'canie inhibited, the cells would not be revealed by our impreg- 
 nation techniciue. Inliiliition of tyrosinase activity by SH groups has been 
 j)roved to occur in adult wiiite skin. In tlu' jialm and sole of Negroes it 
 begins ])erhaps as early as the fourth fetal month. 
 
MELANOBLASTS A\D MELANOCYTES IX FETAL NEdRO SKIN 29 
 
 VII. Individual Variability in the Population Densities 
 of Melanocytes During the Fetal Period 
 
 The poohng of cell counts tluough a relatively extensive develop- 
 mental period, as done in the foregoing section, has the advantage of 
 diminishing the effects of minor technical errors and of bringing out the 
 major trends of conditions under study. Such a procedure has the dis- 
 advantage, however, of erasing iiuUvidual differences in the population 
 density of melanocytes. The "average truth" is not the whole ti'uth in 
 these matters, any more than in many others. 
 
 A possible source of error resides also in relationship between the 
 degree of differentiation of a specimen and its age as determined by exter- 
 nal body dimensions. Streeter showed in his "horizons of human develop- 
 ment" that an identical degree of interior differentiation may be attained 
 in embryos of different external dimensions. During fetal development 
 that discrepancy probably is less significant. Our age determinations ac- 
 cording to Scammon and Calkin's formulae are valid for comjiarisons, but 
 they do not necessarily reflect absolute age. Any errors in this respect 
 probably would affect age determinations by not more than one week. 
 
 Already in the twelfth week we encountered one specimen in which 
 the average count of melanocytes per mm- was much higher (540) than 
 the entire range for other individuals of that period (50-200). A similar 
 difference between specimens of practically identical crown-rump lengths 
 was noted at 15 weeks. The discrepancy in cell counts was especially 
 apparent in corresponding skin areas of the lower limbs. Again in the 
 eighteenth week we obtained range variations in cell counts from 300-700 
 in one individual and from 600-1,000 in another. The range of variability 
 generally was between 500-1,000 melanocytes/mm- in specimens of the 
 nineteenth week. In one individual of the twentieth week the counts 
 varied from 700 to 1.100 cells and in another fetus of the twenty-first week 
 from 500 to 900. In two fetuses of the twenty-sixth week the cell counts 
 (disregarding palm and sole) were between 675 and 1,100 per nmr in one 
 specimen, and between 1,000 and 1,485 in the other. Counts in skin speci- 
 mens of newborn Negroes averaged 1,035 melanocytes per nun-. Unfor- 
 tunately we were not able to obtain material from many cutaneous areas of 
 that age. 
 
 The above-reported individual differences might be considered char- 
 acteristic only for fetal stages. Our data do not extend far into postnatal 
 life and we cannot say whether such differences persist. It is well estab- 
 lished, however, that they exist both in white and Negro adults. 
 
30 ZIMMKHMANN AND HKCKKli 
 
 VIII. Discussion 
 
 111 the ])ast, cutaiioous pifiincnt cells and their precursor stapes have 
 been known !)>■ \'ari()us names. To the jn'oponents of a theory that preceded 
 Hloch's. tlu\\- were tixcd conncri ivc tissue cells that had penetrated the 
 epidermis. Accordinji to Mlirmann. pigiiieiit-i)roducinf;- cells were "melaiio- 
 blasts," located in the dermis or in the epidermis and presumed to be 
 derived from embryonic mesoderm. But Bloch insisted that they were 
 of ectodermal origin, arising in the basal layer of the e]Mdennis. Biologists 
 generally jireferred the terms chromatophores or melano])liores. evidently 
 disregarding the fundamental difference between carrying melanin and 
 actively producing it. 
 
 Adachi, with other distinguished authors of his time, believed that 
 "epidermal chromatophores'" were an illusion. ^Melanin of the mammalian 
 and human epidermis was not contained in cells at all but in intercellular 
 spaces. Others, i)articularly P>ench authors, designated the epidermal 
 pigmentary elements as "Langerhans cells." Paul Langerhans (1868) had 
 discovered stellate cell forms in adult white skin by means of gold imjireg- 
 iiations. He believed them to be nerve cells and did not associate them 
 with pigment ]iroduction. Later these elements were considered either as 
 artifacts oi- identical with the true ]5igment-prii(lucing dendritic cells of the 
 epidermis. 
 
 According to Masson (1!)48), "Langerhans cells" represent effete 
 melanocytes that have lost their ability to produce melanin and are 
 ajiproaching desciuamation. Billingham and Medawar (1953) emphasized 
 tliat only "high level" branched cells in the epidermis are identical with 
 "Langerhans cells" and that the latter did not occur in the basal layer. In 
 adult heavily pigmented skin the high level dendritic cells never are ilojxi 
 positive. They have either lost or discharged their pigment and are an 
 exhausted type of cell. 
 
 In fetal Negro skin no distinction can be made between branchetl cells 
 in superficial or deep layers of the epiilermis. All are functionally active as 
 pigment producers. Consequently, there are no "Langerhans cells" in 
 fetal Negro skin. This confirms the interpretation that in adult skin they 
 represent "spent" melanocytes. Hence we have not usetl the term. 
 
 Masson (1948) designated the pigment-producing elements of the 
 epidermis as "clear cells" or "cellules claires." Tlie term staiuls for cell 
 bodies (perikarya) that lie mostly in the basal layer and usually show 
 some pigmentary activity. They were even described as lymphocytes 
 penetrating the epidermis. Billingham and Medawar (19.13) ha\e drawn 
 attention to that erroneous interpretation. 
 
 liloch (1917) and Beckei- (1927) called the dopa-]iosili\e biancluMl 
 cells I)( iiilrlh iizi ll( II nv (lench'ilic cells. These leiins were ])arl icularly 
 
MELAXOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 31 
 
 useful in the earlier days of pigment research when neither their origin nor 
 true nature was known. "Dendritic cell" now is a widely used term 
 (Billingham. 1948). It is descriptive, noncommittal, and applicable to all 
 epidermal stellate cells, whether they have pigmentary activity or not. 
 
 In this bewililering array of more or less synonymous terms — which 
 naturally has led to confusion — the best name for pigment cells is "mel- 
 anocytes." We have applied it both to immature, branched forms of 
 pigment-producing cells in the fetal dermis and to the highly differentiated 
 dendritic cells of the epidermis. Pigmentary as well as inhibited dendritic 
 cells are melanocytes. Alelanoblasts, instead, are early embryonic cell 
 types, round or ovoid, and re\'eal the first signs of their j^igment-producing 
 potentiality. 
 
 During the past 75 years a bi'oad concept has evolved relative to 
 melanocytes in the human dermis. At first recognized only in the sacral 
 region of Japanese fetuses and children, then in the dermis of certain 
 primates, the same cells were subsequently described in the sacral region 
 of whites, in blue nevi, and finally in the dermis of several body regions of 
 early fetuses of the yellow race. We have shown that during early fetal 
 months, immature melanocytes are present in the dermis of all body 
 regions of Negro fetuses. j\Ieasurements of the characteristic cells corre- 
 spond closely to those given by previous authors for the so-called Mon- 
 golian cells. Barry (1953) postulated that "dermal melanocytes are the 
 forerunners of dendritic Langerhans cells." He fuither emphasized that 
 the pigment cells of the Mongolian spot, of the blue nevi, of the hair 
 matrix, and the dendritic cells of the epideimis are cytogenetically related 
 to each other. In general, we agree with this conclusion. 
 
 We have found that in later fetal life and at biith, "dermal" melano- 
 cytes remain identifiable only in certain skin areas: in the scalp, the sacral 
 region, and the dorsum of the hand and foot. Temporarily, all cutaneous 
 areas of Xegro fetuses contain some pignient-j^roducing cells in the dermis. 
 These melanocytes appear in every way homologous to the permanently 
 present pigment cells in the dermis of certain anthropoid apes and 
 monkeys. 
 
 Their gradual disappearance from the dermis of most skin areas needs 
 further investigation. The present methods did not permit us to establish 
 whether many of the melanocytes degenerate or whether their enzyme 
 system merely becomes inactive. Possibly some of the cells ascend to the 
 dermo-epidermal junction and mature into highly branched epidermal 
 melanocytes. 
 
 Figure 2 shows our interpretation of the paths followed by melano- 
 blasts and melanocytes. Since melanoblasts were not seen before the tenth 
 week of menstrual age, parts A and B of the diagram remain hypothetical. 
 
32 ZIMMKHMANX AND HKCKKIi 
 
 yt 
 
 ■p 
 
 ', ,. 
 
 1, 
 
 1 
 
 
 7^ 
 
 r' 
 
 Q 
 
 
 1. 
 
 
 «i 
 
 c «■ 
 
 -^ 
 
 ^ 
 
 ^ 
 
 5 
 
 5 "S ■ 
 
 J 
 
 -^ 
 
 6 
 
 melanin c 
 eos (Mong> 
 Blue Nevi 
 
 ;- 
 
 X 
 
 
 ^ ^ ^ 
 
 
 
 ^ 
 
 005 
 
 
 
 ^ 
 
 c ;s 
 
 E 
 
 ^ 
 
 X 
 ^ 
 
 X 
 
 = 
 
 1 
 
 ^■5 1 
 
 «, <o C 
 
 
 ^ 
 
 Vl Q 
 
 >, 
 
 
 ^ 
 
 d) ■*? c 
 
 
 
 ^ 
 
 S-s-§ 
 
 ;i 
 
 •I! 
 
 1 
 
 
 
 
 
 «j »^ 
 
 
 
 1 
 
 1! 
 
 ^ 
 
 ^ 
 
 ^<b 
 
 Q 
 
 ^ 
 
 '^s 
 
 ^ 
 
 
 
 rH 
 
 -♦-» 
 
 s 
 
 a 
 
 Si 
 
 
 
 xrt 
 
 « 
 
 ■tj 
 
 1 
 
 
 <J 
 
 % 
 
 g 
 
 
 
 
 
 c 
 
 ;-l 
 
 S 
 
 tt-l 
 
 
 
 "3 
 
 d. 
 
 p 
 
 
 
 
 -*J 
 
 
 !C 
 
 ^4-4 
 
 
 
 
 
 
 
 ^ 
 
 c 
 
 -^ 
 
 .2 
 
 
 '-H 
 
 c 
 
 .S 
 
 C3 
 
 
 r-* 
 
 '-+3 
 
 e 
 
 O o 
 
 '^l 
 
 C 
 
 a> 
 
 ^ 
 
 tt. 
 
 a> 
 
 
 ■a 
 
 
 
 
 
 
 
 
 
 S 
 
 ■^ 
 
 
 
 
 
 
 en 
 
 CA: ■- 
 
 <N .S t. 
 
 
MELANOBLASTS AXD MELANOCYTES IN FETAL NEGRO SKIN 33 
 
 They are based on experimental evidence in mammals (Rawles, 1947). 
 The remainder of the schematic representation is self-explanatory. The 
 cell types shown were dra\\-n from high po^^•er observations. The evidence 
 for our views, of course, is mainly circumstantial since actual cell migra- 
 tions and transformations cannot be observed in fixed tissues. This is a 
 corollary of a purely morphological or histological approach. 
 
 Another question pertains to the multiplication rates of melanoblasts 
 and melanocytes in their migratory phase. Hu, et al. (1957) have shown 
 that melanocytes in cultured sheets of Xegro epidermis underwent mitotic 
 divisions. Billingham and Sparrow (quoted by Billingham and Aledawar, 
 1953) have proved experimentally that the pigmented epidermis is "a 
 reproductively self-sufficient system." Cell divisions occurred in both ^lal- 
 pighian cells and melanocytes. Masson (1948) was convinced that "mel- 
 anoblasts" multiply within the basal layer of the human epidermis. He 
 noticed features suggesting amitotic division and only once recognized a 
 mitotic figure. Cell divisions of epidermal melanocytes have also been 
 reported by Pinkus (1949) in a wart and by Becker Jr., Fitzpatrick, and 
 Montgomery (1952). 
 
 We were unable to study nuclear events, since the silver impregnations 
 of premelanin and melanin granules largely obscure them. Many cell 
 forms of immature melanocytes in the dermis recalled Masson's observa- 
 tion. The manner of their possible multiplication needs further study. The 
 numerous round melanoblasts in the scalp in late fetal periods and even 
 at birth are also puzzling. Both cell forms may represent preparatory 
 phases of cell division, ratliei' tlian new cells derived from the neural crest 
 (Weissenfels, 1956). Other staining methods will have to be employed to 
 obtain a flefinite solution to this problem. 
 
 The main data on population densities of melanocytes in various body 
 regions and at different times of the fetal period are briefly summarized in 
 the conclusions of the present study. 
 
 Previous studies have shown that the average number of epidermal 
 melanocytes per unit area is not significantly different in adult individuals 
 of the white and Xegro races. ^Marked numerical differences, however, do 
 exist between individuals of either race (Szabo, 1954; Staricco and Pinkus, 
 1957). We found similar individual variations to exist already in the 
 fetal period of X'egroes. Our data indicate that differences in population 
 densities of melanocytes, both regional and individual, probably become 
 established relatively early in fetal life. 
 
34 zi.MMi:i(MA.\.\ AM) ni;(Ki;u 
 
 IX. Summary 
 
 'I'hiii lioiizoiital sli;uiiiji.s of tVtal Xi'jiiu t^kiii were iinprejiiiated with 
 ainiiioiiiiical silver iiiliatc. A inodificfl Masson tcchiiiciuo was used and the 
 split-skin picpaiatioiis wcic iiKnintcd as spreads. The material consisted 
 of 1()() fetuses, varyiiifi in ajje from the seventh week to birth. Twenty 
 selected skin areas were studies in each fetus. Over 500 field counts of 
 pigment cells in the dermis and epidermis were made. 
 
 The earliest precursor stages were idcntifiiMl in the dermis of the tenth 
 week of development. They appeared as round nielanoblasts containing 
 the first argentaffin granules. Still younger forms of that embryonic cell 
 type could iiot be distinguished from mesenchymal cells. A gap of a few 
 weeks remains between the development of the neural crest in man and 
 the earliest identification of its presumptive pigmentary elements. Experi- 
 mental evidence in mammals has closed that gap (Rawles). 
 
 Between the tenth and twelfth weeks, nielanoblasts occur in the dermis 
 in increasing numbers; the epidermis then contains only scattered den- 
 dritic cells. Mclanoblasts dift'erentiate into spindle-shaped, immature 
 melanocytes with two or three short, stubby processes and increasing 
 amounts of pi'cmelanin granules. They were identified in the dermis of all 
 body regions, in groujjs or "swarms." Cell counts were highh' A^ariable. 
 
 Immature melanocytes were recognized in unstained preparations as 
 early as the fifteenth week. Average counts of 300 to 400 per mnr' were 
 obtained in the thirteenth and fourteenth weeks, of 500 to 700 in the 
 fifteenth and sixteenth weeks, and over 1,000 in the sacral region from the 
 seventeenth to the twentieth week. From the sixth month to birth, 
 "dermal" melanocytes were found only in the scalp, sacral area, and in the 
 dorsum of hand and foot. 
 
 The first few epidermal melanocytes were observed in the eleventh 
 week. Their numbers per mm- increased sharply between the twelfth and 
 fourteenth weeks, indicating a ])eriod of rapid infiux. No transitional 
 forms of ordinary basal cells were seen. At first fusifoi'm with long pri- 
 mary ])rocesses, the epidermal melanocytes rapidlj' diti'erentiated into large 
 dendritic cells. As early as the twelfth week they appeared in most body 
 regions, including palm and sole. Uniform distribution patterns and high 
 poiiulation densities became established before the end of the fourth 
 month (SOO to 1,000 melanocytes jier nun-'). Their number decreasetl in 
 palm and sole with the deNeluiiiiicnt of rele ridg(>s (sixteiMith to nine- 
 teenth weeks), but some melanocytes remain on the ridges until i)irth. 
 Population densities of ejiideiinal melanocytes remained fairly stabilized 
 after the fifth fetal nmnlh. in the newborn Negro there were approxi- 
 mately' l.Oiv") dopa-])osil i\e dendritic cells per iinn-'. 
 
 Regional and in(li\idual dilTerences in poimlation densities of fetal 
 
MELANOBLASTS AND MELANOCYTES IN FETAL NEGRO SKIN 35 
 
 melanocytes were observed. An early cephalo-caudal gradient later dis- 
 appeareil. It probably expressed different times of "arrival" of epidermal 
 melanocytes in various body regions. The earliest fully elaborated melanin 
 first appeared in certain head regions (third month). Gradually a dorso- 
 ventral gradient in population density of epitlermal melanocytes emerged 
 and remained until birth, especially in the trunk and upper limb regions. 
 Individual differences in jiopulation densities were noticed as early 
 as the fourth fetal month. 
 
30 ZIMMKIIMANN AM) Hi;( KKK 
 
 X. Literature Cited 
 
 AiiA< 111. \'>. IW.'i. ll:iui|iifiiiifiit In'iiii .Mi-nscla'ii uiid Ijfi ik'ii AftVn. Zcilsclir. f. 
 
 Mdipli. u. .\ntlii(i|i. li: 1-131. 
 Baklz. K. 1885. Die krupcrliclicn Eisicnscliafteii (Icr .lapaiK r. Mittcilp. DcTitst-li. 
 
 (it'.-:i'li.-^cli. f. Natur-u. \'olkc'rkun(it' O^-tasion.s. 4:40. 
 Bakuv. .1. 1953. Uccliirclics sur I'originc dcs cellules de Langerhans de repiderme. 
 
 Biol. med. 4,-': 293 -31.",. 
 Bkckkr, S. W. 1927. Mthiiiin iiiiiiiuiitaiiini. A .-^y.-tiiuatic .-^tiuly of tiie ]iigincnt 
 
 (if tlic iiuiiiaii skill and upper iiiucous iiieiiibranci?. with P]iccial consideration 
 
 of pigmented dendritic cclL-^. Arch, of Derinat. and Syi)h. J'j:259-290. 
 Bkckeu, S. W., Jr., Th. B. Fitzp.\trick. and H. Mo.nt(;omery. 1952. Human 
 
 inelanogcnesis — Cytology and histology of jiigment cells. A.M. A. Arch, of 
 
 Dermat. and Sypli. 6V;:511-523. 
 Becker, S. W., Jr., and A. A. Zi.m.mkk.maxn. 1955. Further stuilies on melano- 
 cytes and mclanogcncsis in the human fetus and newborn. J. Invest. Dermat. 
 
 ;85: 103-112. 
 BiLLiNCHAM, R. E. 1948. Dendritic cells. J. Anat, ,9.^:93-109. 
 
 . 1949. Dendritic cells in jiigniented human skin. J. Anat. 55:109-115. 
 
 and P. B. AIedawar. 1953. A study of the branched cells of the mam- 
 malian epidermis with special references to the fate of their division products. 
 
 Phil. Trans. Roy. Soc. London, Scries B, ^57:151-171. 
 Bloch, B. 1901. Preuvcs atavi(iucs de la (ransfonuatiim des races. Bull. Soc. 
 
 Anthrop. Paris, V Series. ;?:618. 
 . 1917. Das Problem der Pigiuenlliildung in der Haul. Arch. f. Dermat. u. 
 
 Sypli. i,^4: 129-208. 
 . 1921. Leber die Entwicklung des Haut — und Haarpigmcntcs beim 
 
 menschlichcn Embryo und ueljer das Erloeschcn der Pignientbildung im er- 
 
 grauenden Haar. Arch. f. Dt'rniat. u. Syph. 755:77-108. 
 Dannkkl, R., and G. ("i.ekk.maxx. 1954. Die Einwanderunsi der Pigmcntzcllcn 
 
 in die Haut und die Haarc bei Nagetieren. Biol. Zentralbl. ?.>:414-428. 
 EiiHM.WN, S. 1885. Lntersucliungen ut'bcr die Physiologic und Pathologic des 
 
 Hautpignientcs. .\rch. f. Dermat. u. Syph. 77:507-532. 
 Ei, Bahhawv. .v. ,\. 1922. Leber den Mongolenfieck \n'\ iMiropilirii. Arch. f. 
 
 DeniKil. u. Sypii. 74/: 171-192. 
 EsciiiiiciiT, D. F. 1849. Zool.-aiiat.-jiliysiol. Liitersiuliungen iibi'i- die nor- 
 
 disciieii Walltiere. 7:70. Ti. \'oss, Leipzig. 
 (luiM.M, F. 1895. Beitriigc zum Studiiiiii des Pigiiieiits. Dermat. Zeitschr. i:328- 
 
 343. 
 Iliii'Ki.\s I'ux. .\l. 1949. .\iial\>is of smiie phases of iiirlaiiolil.-ist migration in the 
 
 l)arred I'lyiiioulli Rock embryos. Physiol. Zool. ,^,^:l-22. 
 Ilr. I''., R. Iv SxAiutTo. II. PiXKi s. and R. P. FosNAiiiii. 1957. Human iiicl.ano- 
 
 i\tes in tissue culture. .1. liu'est. Dermat. As": 15-32. 
 IsiiiKAw.v, X. 1924. Leiier den sogcuannten Mtmgolenfleck l)ei Japanisclicn 
 
 Foett'ii. Folia anat. japouica, i:l-4. 
 Pro, M. 1953. Stiulies on Melanin. Tohoku .1. I'.xper. .Med. •'i i, Suppl. 1. 
 
MELANOBLASTS AND MELANOCYTES IX FETAL XEGRO SKIX 3/ 
 
 . 1957. Studies on Melanin. Report II. Tuhokii J. Exper. Med. 65. Siippl. V. 
 
 K.\TO, T. 1905. Anatomisch-liiftologische Stiidien iiber die sog. Kinderflcckc. 
 
 Mitteilg. :Med. Fakult. Univ. Tokyo tf: 377-396. 
 Laxgherhaxs. p. 1868. Ueber die Xerven dor Haiit. Virchow's Arch. 44:3'2o-337. 
 Massox, p. 1948. Pigment cells in man. in 'Biology of ^lelanoma?," Spec. Publ. 
 
 X. Y. Acad. Sci., 6:15-51. 
 ^IiESCHER. G. 1922. Die Chromatophorcn in der Haut des Mensclien. Arch. f. 
 
 Dennat. u. Sypli. ;o'i :313-425. 
 MoRisox. Dr. 1889. Beitriige zur Frage von der Pigmentbildung in der Neger- 
 
 haut. Monatsheftc f. prakt. Dennat. ,9:485-490. 
 PixKUS. H. 1949. Mitotic division of lunnan denilritic mclanoblasts. J. Invest. 
 
 Dermat. 75:309-311. 
 Rawles, M. E. 1947. Origin of pigment cells from the neural crest in the mouse 
 
 embryo. Physiol. Zool. 20:248-266. 
 . 1948. Origin of melanophorcs in development of color patterns in verte- 
 brates. Physiol. Reviews, A'^: 382-408. 
 ScAMMOx, R. E.. and L. A. Calkins. 1929. The tlevek)i)ment and growth of the 
 
 external dimensions of the human body in the fetal period. University of Min- 
 nesota Press. 
 St.\ricco, R. J., and H. Pixkis. 1957. (.Quantitative and qualitative data on the 
 
 pigment cells of adult human epidermis. J. Invest. Dermat. 25:33-45. 
 Szabo, G. 1954. The numi)er of melanocytes in lunnan eiiidermis. Biit. Med. 
 
 Jour. ;: 1016-1020. 
 Thomsox, A. 1891. Note on the skin and seal]) of the Negro fetus. J. Anat. 
 
 Physiol. i.J: 282-285. 
 Weissexfels, X. 1956. Licht-phasenkontrast-und elektronenmikniskoiiische Un- 
 
 tersuchungen iiber die Entstehung der Propigment-Granuia in Melanoblasten- 
 
 kulturen, Zeitschr. f. Zellforsch. 45:60-73. 
 Zimmer.mann. a. a. 1950. The development of epidermal pigmentation in the 
 
 Xegro fetus. Proc. 2nd conf. on biology of norm, and aty])ical pigment ceil 
 
 growth. Zoologica. 55:10-12. 
 . 1954. Die Entwicklung tier Hautfarbe beim Xeger vor der (ieburt. 
 
 ]Mitteilg. Thurg. Xaturforsch. Gesellsch. (Switzcrlanili 57:33-71. 
 and Th. Corxbleet. 1948. The develoimient of epidermal |iigmentation 
 
 in the Xegro fetus. J. Invest. Dermat. ;;:383-392. 
 
PLATES 
 
 Figures 3 to 26 are surface views of eitlicr full- 
 thickness or split-skin preparations of Negro fetuses. 
 All except that shown by figure 7 were impregnated 
 with reduced silver by a modified Masson technique. 
 Figure 7 represents an unstained preparation. The 
 magnification varies between 700 and 800 X . 
 
 39 
 
PLATE 1 
 
 Furt'aiiii. 12 wciks. Full-thickness skin spreads. 
 Early precursur ^tai^c^: oi nu'lanocytcs in the dermis. 
 
 3 Round nu'lanublast, 8-10 /i diauR'tcr. 
 
 4 Early iiinnature niclaiidcytt' t'drniing three ])roc- 
 esscs. 12-16/1. 
 
 5 Early fusilorni iiiiniature melanocyte, 20-25 {i. 
 
 6 Highly fusiform immature melanocyte 40-50 /i. 
 
 40 
 
;SPf;- 
 
 'jw^ 
 
 41 
 
PLATE 2 
 
 7 Scalp, 15 weeks. Uiistaiiu'il. Two spindle-shaped, 
 iiuniaturc melanocytes in the dermis. A small hldud 
 vessel is coursing between them. 
 
 8 Dorsum of foot, 14 weeks. Fusiform melanocytes in 
 the dennis. 
 
 9 Lvuiibdsacral region, 13 weeks. Fusiform and early 
 stellate forms of iniiiiature melanocytes in the 
 dermis. 
 
 42 
 
43 
 
PLATE 3 
 
 10 Dorsum of liaiul, IG weeks. \'ari(ius forms of im- 
 mature melanocytes. 
 
 11 Sacral region (Mongolian spot I, 19 weeks. Round 
 and fusiform, immature melanocytes. 
 
 44 
 
45 
 
PLATE 4 
 
 12 Scalp, 18 weeks. AlelaiKR'vtes adjacent to a small 
 Ijlood vessel. 
 
 13 Dorsum of hand, full-term fetus. Round, fusiform 
 and stellate types of melanocytes in the dermis. 
 
 4(1 
 
47 
 
PLATK 5 
 
 14 InttTscainilar legion, 10' i; wt'ck.s. Earlici^t oi)iikT- 
 mal melanocyte witli three ])roeesses. Neighboring 
 nuclei are of basal epidermal cells. 
 
 15 Sole, 12 weeks. Some of tlie earliest melanocytes 
 appealing in the e])i(leniiis are highly fusiform, 
 usually showing two long processes. 
 
 16 Nape, 12 weeks. Early epidermal melanocytes 
 with 2-3 dendritic ])rocesses. 
 
 48 
 
49 
 
PLATE 6 
 
 17 Scalii. 12 weeks. Earliest epiileniial melanocytes in 
 small groups, irregularly scattered tludugh the epi- 
 dcnnis. Cell counts are unreliable. 
 
 18 Anterior region of leg, 12 weeks. Large, heavily 
 stained melanocytes appear here and there. The 
 majority of epidermal melanocytes are smaller and 
 relatively uniform in size. Neighboring nuclei rep- 
 resent basal cells. 
 
 50 
 
51 
 
PLATE 7 
 
 19 Nape, I2Y2 weeks. Tlic uiuqual degree of silver 
 impregnation of various melanocytes probably re- 
 flects uncfiual maturity or functional activity. 
 
 20 Dorsum of arm, 12 weeks. Fusiform anil stellate, 
 early dendritic cells. 
 
 52 
 

 53 
 
PLATE 8 
 
 KifliiT distribution i);itl(.Tiis of I'piilfiiiial uiflaiio- 
 cytos. 
 
 21 Scalp, 13^0 weeks. Regular distribution and fairly 
 even size of ciiidcrmal melanocytes in an intcrfol- 
 licular field. 
 
 22 Interscapular reiiion, 1.5 weeks. The luelanoeytes 
 appear to be oriented with their long axes ruiniing 
 parallel to each other. 
 
 54 
 
55 
 
PLATE 9 
 
 23 Palm, 13' j weeks. Fairly i'\'en ilistrihiitidii of ei>i- 
 dei'iiial iiiclanoeytes hi'lore retc ritlges (ei)i(lcrnial 
 crests I are dc^'i'liiiuMJ. 
 
 24 Palm, 17 weeks, ^^■itll the development of rcte 
 ridges the melanocytes come to lie on the crests. 
 Note the de\elo|)inir sweat pores. 
 
 56 
 
■ --^ ,- ■'■■ 
 
 ik 
 
 
 
 
 23 
 
 57 
 
PLATE 10 
 
 25 Anterior ahildiiiinal wall, 19 wuek?. Rich pattern 
 of intracpidt'rmal dendritic cells. 
 
 26 Palm, newborn. Epidermal melanocytes and sweat 
 pores are recognizable on the rcte ridges. 
 
 58 
 
^ 
 
 
 w 
 
 'S 
 
 --10: 
 
 \ .0 y 
 
 59 
 
ILLINOIS MONOGRAPHS IN THE MEDICAL SCIENCES 
 
 Vol.1 
 
 No. 1. The Magdalenian Skeleton from Cap-Blanc in the Field Museum of 
 Natural History. By Gerhaedt von Bonin. $1.00. 
 
 No. 2. Studies in Infant Speech and Thought. Part I. The Development of Sen- 
 tence Structure in Infancy from the Viewpoint of Grammar. A Quanti- 
 tative Analysis of the Continuous Speech Record of Two Infants. By 
 Abraham A. Low. $L00. 
 
 No. 3. Classification of Yeasts and Yeast-like Fxmgi. Distribution of Myco- 
 logical Flora on Normal Skin and Mucous Membranes. By C. Virginia 
 Fisher and Lloyd Arnold. $1.00. 
 
 No. 4. Hemophilia — Clinical and Genetic Aspects. By Carroll L. BntCH. $2.00. 
 
 Vol. II 
 
 No. 1. Origin and Development of the Mediastinal and Aortic Thyroids and the 
 Periaortic Fat Bodies. By Otto Frederic Kampmeier. $1.50. 
 
 No. 2. The General Tissue and Humoral Response to an Avirulent Tubercle 
 Bacillus. By Sol Roy Rosenthal. $2.50. 
 
 No. 3. The Vasomotor System in Anoxia and Asphyxia. By Ebnst Gellhohn 
 and Edward H. Lambert. $1.00. 
 
 No. 4. Intestinal Motility of the Dog and Man. By Charles B. Puestow. $1.00. 
 
 Vol. Ill 
 
 Nos. 1-2. Origin and Development of the Lymphatic System in the Opossum. 
 By Arnold A. Zimmermann. $2.50. 
 
 Nos. 3-4. Vascular Spasm — Experimental Studies. By Alexander John Nedzel. 
 $2.25. 
 
 Vol. IV 
 
 Nos. 1-4. The Precentral Motor Cortex. By Paul C. Bxjcy, Editor, and others. 
 2nd edition. Cloth. $6.00. 
 
 Vol. V 
 
 Nos. 1-2. The Reticulo-Endothelial System in Sulfonamide Activity. By Frank 
 
 Thomas Maher. $2.50. 
 No. 3. The Cortex of Galago. Its Relation to the Pattern of the Primate Cortex. 
 
 By Gerhakdt von Bonin. $1.50. 
 No. 4. The Neocortex of Macaca mulatta. By Gerhardt von Bonin and 
 
 Pebcival Bailey. $3.00. 
 
 Vol. VI 
 
 Nos. 1-2. The Isocortex of Man. By Percival Bailey and Gerhardt von Bonin. 
 $5.00.