QUANTITATIVE: STUDIHS ON THE GROWTH O F THE E PIDERSITS LEO LOER A N D FRANCES L. HAVEN Department of Pathology, Washirrgton University School of Medicine, Xt. L O W ~Missouri S. llie aim underlying the following investigations was two- r i fold: In the first place, we wished to obtain data concerniiig the relations between structure and cell proliferation in dif fercnt types of epidermis which would supplement previous ones obtained by Loeb and Addison(1 ), Loeb(2), Addison and Loeb( 3), Spaill (4),Spaiii and Loeb ( 5 ) , and X kaiwa (6). 111 these earlier investigations the authors had shown that there exists a definite relation between structure of the epidermis and its proliferative activity, and that the latter is the primary factor which determines the structure. They conclnded, furthermore, that the proliferative acativity is not of a regeiierative type destined to make good the losses through the degenerative process of keratinization, hut that it is a primary process, apparently fixed hereditarily. They also found that there is a definite relation between the structure and the proliferative activity of tlie normal epidermis and the intensity with which regeneration takes place if R defect has been made. In the second place, we wished to find a method which would make it possible to study cyuantitative1-y the effect of various conditions, inner as well as outer, on the proliferative energy of a certain tissue, and, furthermore, to provide the basic data concerning the normal activity of the tissue, which could be used as a standard; the deviations from this standard could then be attributed to the influence of the variation of factors in the inner or outer environment of this tissue. As the most convenient tissue we selected the epidermis, 21 7 'If18 ANATOXTCAL R E r O l t l ) , Y O L . 42, NO. 3 MAY. 1929 218 LEO LOEH A N D FRANCES L. H A V E S primarily that of tlie car of the guinea-pig; hut we compared this also with the epidermis of other animals and with the other parts of the skin of the guinea-pig. We hoped thus to provide a basis for a quantitative study of the factors underlying abnormal growth processes in the body, in addition to aiialyzing the variations coming within the range of tlie normal. In order to obtain comparable data of the proliferative activity of the epidermis, we determined the following factors : 1. The iiuml)er of mitoses per 1 mm. skin. 2 . The 1iuml)er of mitoses per 1000 cells. 3. Thc nnmher of mitoses in tlic lower layer of epidermis ~ ) e 1000 r cells. 4. The iinmher of mitoses iii the lower layer of tlie epidermis per 1 mm. skin. 5. The number of mitoses in the upper layers of the epidermis per 1 mm. skin. 6. The total number of not yet keratinized cells per 1mm. epidermis. '7. Thc number of cells in tlie lower layer of the epidermis per 1mm. skin. 8. The number of cells in the upper layers per 1 mm. skin. We made counts in R numhcr of individual animals and then took the averages. As f a r as the number of mitoses is concerned, we consider as the most important figure the number of mitoses per 1mm. skin. I n the same type of skin this figure depends on a smaller numbcr of variables than the number of mitoses per 1000 cells; accordingly, whereas the number of cells in the lower layer is very constant in the same type of skin, thcre is a considerable variation in the number of cells in the upper layers under differen t conditions. The greater the number of cells, the smaller will be the number of mitoses per 1000 cells, and vice versa. Because of the relative constancy of the number of cells in the lower layer of the epidermis in the same type of skin and because of the fact that the great GROWTH O F T H E EPI1)EItMIS 219 majority of mitoses are in the lower layer, the number of mitoses per 1000 cells in the lower layer is a much more constant figure than the number of mitoses per 1000 total cells. In the same type of skin the number of mitoses in the lower layer per 1000 cells is similar to the number of mitoses here per 1mm. skin; it is therefore possible to omit consideration of the number of mitoses in the lower layer, if we determine the number of mitoses per 1 mm. skin. However, in a number of the more important conditions the average for the number of mitoses in the lower layer m-as determined in adnures. ditiori to the other fig Before beginning an analysis of the factors underlying growth processes in the epidermis, it was necessary to determine the degree of accuracy of our methods. For this purpose we repeated the countings made in a piece of skin and noted the variation in the two counts. We next used skin from both ears of the same animal, in order to compare the variations in counts thus obtained with those obtained when ears of different animals were used. EAR SKIN O F GUINEA-PIG Variations in coum5n.g.s in the same a!izimal as a test of flw degree of accuracy of the method used by tis I. Double counts. Nine guinea-pigs. The same piece of ear skin was counted, but the areas counted in both cascs were presumably not identical. The mean of total mitoses per 1mm. skin was 3.52 in the first count; in the second count it was 3.88. The percentage difference between the total of first and second counts is 9.27. If we figure the mitoses per 1000 cells instead of per 1 mm., the figures for the first count were 7.19 and for the second count, 7.70. This represents an average percentage difference of 7.22. If, instead of considering the percentage difference between the two counts, we take the actual difference in the number of mitoses counted, this was 0.36 mitoses per 1 mm., and per 1000 cells it was 0.56 mitoses. 220 LEO LOIIiH ANTI FILANCES L. H A V E N In regard to the total number of cells, the average in the first count is 486; in the second count, 475, which is a pcrccntage difference of 2.26. The average of the first count of [*ells in the lowest row is 155; of the second count, 153--a percentage difference of 1.29. Tlie corresponding figures for the upper cell rows are 330 and 522-a percentage difference of 2.42. The greatest divci*gence betwecii first and sccond counts, :is f a r a s number of mitoses is conceriied, was 35 per cent. If, instead of comparing the averages of the first counts and of the second counts of the mitoses per 1000 cells with each other, we compared averages of the lower counts with the avtwiges of the higher counts in this series, the percentage c l i f f c r e ~of~ the ~ averages was slightly higher, viz., 10.38. We may conclucic, then, that this method is sufficic~ntly wcurat e 1o allow definite conclusions if the differences f o m d exceed considerahly those obtained in the double counts. IZ. Comparison of skin cocstfs of both m r s . Our second test f o r the range of error implied in this method consisted in making counts of cells and mitoses from the skin of both ears of guinea-pig o r rat. This procedure promised, in addition, to throw light on the question as to how f a r different parts of the same kind of' tissue, within the same organism, differ in proliferative activity and as to whether under iiormal conditions this proliferation is largely determined by local factors o r by factors which apply generally. u. Counts in guinea-pigs. We determined in the usual way the number of mitoses and of cells, first in the right and then in thc left ear, and we used f o r this purpose nineteen guineapigs. The average of the total number of mitoses per 1 mm. skin in the right ears w m 1.20 and in the left cars, 1.02. This corrwponds to a percentage difference of 14.99 per ceiit, as (*omparedwith 9.27 per cent in the double count in oiie ear. The averages f o r mitoses per 1000 cells were in the right cars, 2.65, i n the left ears, 2.20--a percentage difference of 16.98 per cent, as compared with 7.22 per cent in the doiible counts. The actual difference in mitoses conntecl in 1 mm. GROWTH O F THE EPTDEILMIS 221 skin in the first and second ears was 0.18, and per 1000 cells it was 0.45. The absolute difference was therefore very small. The averages of the total counts of cells per 1 mm. skin in the right ears was 448 and in the left ears, 4 5 3 n o r responding to a percentage difference of 1.10 per cent. The average of the cells in the lowest row per 1 mm. was 161 in the right and the same in the left ear; in the upper layers the averages mere 287 and 291-a percentage difference of 1.37 per cent. As to the number of the cclls, the percentage differences l e r e in both the double counts of one ear and the counts of both ears very small, and they happened to be smaller in the latter than in the former. However, in the case of the mitoses, the percentage differences were grcatcr in the counts of both ears than in the double counts of pieces from the same ear; they amounted to approsimatcly 9 i per cent. Furthermore, the differences between the counts of both ears and the double counts of one ear are still greater if we compare in these cases the averages of the highest counts of mitoses per 1000 cells with those of the lowest; in the former we find a percentage difference of 34.69; in the latter, 10.28. This means that variations in the number of mitoses in individual cases are greater where the skin is taken from both ears than are found when double counts are made of skin taken from the same ear. If we compare the counts in the right and left ears, the higher count may, in one animal, be in the left, and in another, in the right ear. Thus individual differences may be equalized to some extent. These figures then indicate the range of variability in the proliferative activity in the skin of different ears. b. Counts in rats, I n four animals the number of mitoses in 1mm. skin and per 1000 cells and the number of cells were compared in skin from both ears. The average number of mitoses in 1mm. skin in the right ear was 0.69 and in the left ear, 0.52-corresponding to a percentage difference of 24.63. The average per 1000 cells was 3.25 in the right, and 2.50 in the left, ear-eorresponding to a percentage difference of 23.07. The absolute differences in the number of mitoses 222 LEO LOER A N D FRANCES L. HAVEN counted were 0.17 per 1 mm. skin and 0.75 per 1000 cellsdifferences of a similar order to those found in the case of the guinea-pig. If we compare the averages in the ears with the higher number of mitoses, per 1000 cells, with the averages in the ears with the lower number of mitoses, we find a slightly higher percentage difference, namely, 27.32. The average number of cells in tlie lowest row in the right ears was 129 and in the left ears, 127-a percentage difference of 1.55. The average number for the cells in the upper rows was the same in right and left ears, namely, 88. If we compare the counts in the ears with the higher number of cells wiih those in the ears with the lower number, we find somewhat higher percentage differences, namely, 3.12 for the c~bllsof the lowest row and 5.55 f o r the cells of the upper layers. The percentage differences in the rat are slightly higher tlian in tlie guinea-pig, probably because the numlier of rats used is smaller and also because the figures for mitoses arid cells are absolutely smaller in tlie rat ; therefore, differences wliich are observed become relatively more important. 011 the whole, we see in both species a fairly good agreement between the proliferative activity in the right and left cars of an individual animal. I I l . A com)pariso?h between the skin of the ear and chest o f the gzlirmz-pig arid rat a ~ td h chest ~ slc.i*t of the pigeori and c k i c k m . While our quantitative studies of cell proliferation in the skin under various conditions are based mainly 011 the findings in the ear of the guinea-pig, it was of interest to determine how f a r the ear skin differs in its characteristics from other types of skin. I n six guinea-pigs we comptired, therefore, the car skin with the skin of the chest. The differences are quite marked. I n these six animals the average for the number of mitoses per 1mm. of ear skin was 2.63 ; for chest skin, 0.40; the corresponding figures pcr 1000 cells are for ear skin, 5.49, and for chest skin, 1.80. The average of the mitoses per 1000 cells in the lower row is 10.T9 for ear skin and 3.32 for chest skin. The average of the total _- _. d. c. E;c1 ~ , 129 127 217 21.5 ... 88 88 287 "92 330 322 -~ ........... .- . . 156 3'73 118 ' 104 91 129 123 113 169 68 I .in 309 158 . 311 297 160 32.5 139 168 302 1.57 310 313 161 252 158 1.58 260 161 ' 237 1957 2 74 1.6i 1 230 265 160 161 161 1'm 183 -. 448 453 486 2.26 yo 475} differciicr 470 6 222 6 220 12 236 9 197 7 ti7 , 468 .55 469 457 2 21 484 32 460 4 1 I 467 474 14 410 27 22 418 8 398 43 3 14 10 386 42.5 17 4 4 19 19 9 9 _ _ -. . . . . . . . . I Ear I1 Cornprison of Iboth ears in mts: e. E a r 1 f. Ear 11 Coiuparison of ear nnd rhrst skin in guiiiea-pigs g. EaT h. Chest i. Ear skin in rat (16 pieces) .j. Chest skin in rat 1. Chest skin of pigwn :md cliirken nb. Molr adult guiiicn-pigs (ear skin) n. Ear skin adult guinea-pig; summer o. Ear skin adult guinea-pig: winter p. Ear skin adult guinea-pig; intermediate season q. Ear skin of guincvi-pigs rluring year I r. Ear skin of guinea-pigs rluring rear I1 Y. Ear skin of young guincii-pigs t. Ear skin of young guinea-pigs u. Summer 2'. Winter :t11(1 iiitermediate I. Young nialr guinea-pigs y. Young frm;ilr giiinea-pigs Recounts : a. Count I b. Count TI Comparison of both ears in guinea-pigs: - -- CIIARACTKR ( I F AF131.11.6 OR COXVITIOh I N WHICH COI:NTY WEBE MADE %- . .. . 2.63 0.40 0.74 0.81 0.27 2.29 2.40 1.34 2.73 2.2.i 2.47 2.21 1.60 1.71 2.04 1.81 1 . ..... 5.03 3.63 5.04 2.92 3.59 4.80 .i 17 4.69 3.94 4.14 4.81 5.49 1.80 3.42 3.27 1.4.5 1 . . . . . . 0.69 24.63 % 3.25 2! % difference 2.80 differenrt 8.00 12.04 5.39 13.91 11.23 10.79 3.32 3.81 6.3 2.15 . .~ 7.19' 7.22 % 3.32 9.27 yo 3.88}difFerenlcr 7.70 )thlrrrnic-c . 0.Se . .... 224 LEO LOER AND FRANCES 1,. HAVEN iiumhcr. of cells per 1mm. for ear skin is 479; for chest skin, 322; the figures for cells in the lower row are 156 and 118 a r ~ l in tlie upper layer 323 and 104, respectively. The chest skin is thus a very much more inert tissue than the ear skin. The number of mitoses is about 74 per cent lower and the number of total cells about 52 per cent lower than in thc average ear skiii. There is a certain difference also in the iiiimher of cells in the lower row of cells, it being 25 per cent lower i n the chest skin; but the main difference concerns the number of cells in the upper layers as ~ ~a s lthe l number of upper layers. The proliferation being less active in the (ellest skin, the number of cells pushed into the upper layer,.i is correspondingly smaller. The chest skin of the guiiieapig rcsemhlcs more the general rat a n d pigeon skin than the ear skin of the guinea-pig. These characteristics of ear and ciicst skin hold good in tlie case of the arerage, as well a s in t lie cdase of the individual, animals. However, a certain rangy of variability exists in the ear skin as well as in the cliest skin; whilc tlicre is iioticeahlc a certairi parallelism in the variatioiis of these two types of skiii in individual animals, this partillelism is not perfect, arid in some animals the counts in tlie car skin map be relatively high, while the correspoiidiiig counts in the chest skin are relatively low. I n the rat, also, WT compared ear and chest skin. The ear skin iii twelve rats was used and in four of those we examined two pieces, SO that w(1 c*oiintedaltogether sixteen pieces. The average counts were as follows: Total number of cells per 1 mm., 220. Cells in lower layer per 1 mm., 129; in upper layers pcr 1nim., 91. There are altogether two to three rows of cells prcseiit in this skin. Number of mitoses per 1 mm., 0.74; per 1000 cells, 3.42; per 1000 cells of lower layer, 5.81. The cliest skin of nine rats was examined, and the following averages were found: Total number of cells per 1 mm., 236. ('ells in lower layer per 1mm., 123; in upper layer per 1 mm., 113. Number of mitoses per 1 mm., 0.81; per 1000 (TIIS, 3.27 ; per 1000 cclls of lower layer, 6.3. GROWTH 01’ T H E EP1I)EIIMIS 225 These figures prove that, in contrast to the findings in guinea-pigs, where ear skin and chest skin differ greatly in proliferative activity and in number of cells, in the rat the ear and chest skin are very similar and they are also similar to the chest skin of the guinea-pig. In all three of these tissues (ear and chest skin of rat and chest skin of guinea-pig) the number of cells in the lower layer of the epidermis is about 30 to 40 per cent less than in the lower layer of the ear skin of the guinea-pig, while the number of cells in the upper layer is only one-third (or less) as great in the former tissues as in the latter. Furthermore, as stated, in the chest skin of guinea-pigs and in the ear and chest skin of rats, the number of cells in the upper layers is smaller than in the lower layer-the reverse of the condition found in the ear skiii of the guinea-pig. The difference in proliferative activity of the epidermis appears more marked if we compare the number of mitoses per 1 mm. skin, because the number of cells in the lower layer is smaller and the cells are flatter in the chest skin of the guinea-pig and in both kinds of rat skin than in ear skin of the guinea-pig. This difference is also evident if we compare the number of mitoses per 1000 cells, hut is less marked by this method, because, in the ear and chest skin of the rat and the chest skin of the guineapig, 1000 epidermal cells cover a much greater area than the same number of cells in the ear skin of the latter animal. If we compare the figures for the number of mitoses per 1000 cells in the lower layer of skin, the greater activity of the ear skin of the guinea-pig becomes again quite marked, because the large majority of all mitoses are found in the lower layer and because in the guinea-pig the number of non-proliferating upper cells which has to he added t o the cells in the lower layer, if we count the number of mitoses per total 1000 cells, is much greater here than in the other three types of skin. While in the latter diminution in the number of cells-and, especially, of the upper layers of skin-is associated with a lowering in proliferative activity, there is no complete correspondence between the number of cells and the 226 L E O I,OEI< A X D FLL4XCRS L. HAVEN 1111ml~~ of mitoses and thc proliferative acativity ; thus, in thci chest skin of tEic guinea-pig the proliferative activitv is lower tliiiii in tlic rat, altliougli the iinmlxr of wlls is similar i n hoth. There mnst enter tlicn ariothcr factor, in addition to tlic proliferative activity, which tletermincs the immf.)erof c . r . 1 1 ~ arid tlii~s,pi-esinmably, the readiness with which tlic cells in the upper layer undergo the tlegenerative process of lic~r;i t iniza tioii. s t of the guinea-pig illid to the skin Similar to the ~ h ~ ?skin of ear and chest of the rat is the chest skin of the pigeon ant1 the c3liickc~11. 'l'lit~ ririml)cr of cclls i n tlic loiver 1uyt.r is of the same order iii a11 thcsc types of epidermis, h t hi the upper lijycr there are oiily sixty-ciglit cells per 1 mm. in tlic chest skill of pigeon ant1 c*hicbken, as compared with about, 011e hiindrecl cells in thc other tissnes. The proliferative acIivity is here still lower than in the chest sliiii of the guiiicaijig, ilnd, because of the lower nnmher of cclls h i the uppcbr 1a;vcr a i d the relative: prepontlcrance of the lower lt-tper of c.c~lls iii 1000 c~~11s--wl~icl~ is the active one, i l s far 8 s the ~~roliferation of cells is coticeriied-the number of mitoses per 1000 cells is a little more tliari five times greater (1.45) than tlit1 nnmhcr of mitoses 1 mm. skin, which is onl?- 0.27. 111 the other kinds of cpidermis (rat, chest skin of guinea1)ig) tlic iiicrcasc iii the iiiimher of mitoscs per 1000 cells, 21s wmparcd with the mimher of mitoses per 1mm. skin, is less marked. However, 1)wause of tlic rc?latively small number of cells ill the ny)pc!r row iu lJrds, the gain iii mitoses per 1000 cells in thc lo~verlayer is somclwhat less hcrc tlian in rat i111(1 griiiica-pig ellest skins. Wc may then provisionally clistiiignisli two f a ~ t o i -wliicli s clcterminc the iiuml)er of cells c.oiistituting the epitlcrmis. 1) The greater proliferati\-c a<:f ivity of' (wtliin typcs of' cpitlermis ( ~ a rskin of guinea-pig) I(w(1s to a more rapid piisliing of the cclls into the iippcr Iiiyrs. These cells arc thus rcmoved from the soiirco of oxyg:chii aiid foodstuffs, w1tic.h a r e s i q q d i d 1)s- tlicb vessels i n 1 l i c c w n n c ~ c ivc t tissiic? iiii(lerii(!titli the epidermis. A s ii result 01' (Icfici( I I C ~h i t l i ~sii7)ply of o x y g ~ n tl~genc.riltive , proccsst's, p c k r GROWTH O F THE EPIDERMIS 227 leading to keratinization, take place. The rapidity with which these processes set in is determined by the length of time during which the cells are removed from the base of supply and by the distance which separates the cells from the latter. I n the rapidly proliferating ear skin a greater number of cells will be removed from the cutis during a unit of time than in the other types of skin ; the number of non-keratinized cells will therefore be greater in the former. Thus we find, in general, a correspondence between number of cells in the upper layers and proliferative activity. 2) There are, in addition, variations in the number of cells in the upper layers, which do not run parallel to the proliferative activity. These variations presumably depend upon a tendeircy to undergo keratinization which apparently varies in a specific manner in different species. IT7. The proliferative activity of ihc ear S l i k of normal male guimea-pi*qs. Sixty-seven male guinea-pigs were used, none younger than four weeks of age. The average number of total cells per 1 mm. was 468; the number of cells in the lower layer, 159, and in the upper layer, 309. The iiumber of mitoses per 1 mm. skin was 2.29, and the number of mitoses per 1000 cells, 4.81. If we distinguish between mitoses in the lower and upper layers, we have statistics of fifty-five normal male adult guinea-pigs. The number of total cells was 469 ; the number of cells in the lower layer, 158, and in the upper layer, 311. These figures are very similar to those given above for a still larger number of animals. Tlie mitoses per 1mm. of skin were 2.40; 1.89 of these were found in the lower, and 0.51 in the upper, layer. The number of mitoses per 1000 cells was 5.04 and, per 1000 cells of the lower layer, 12.04. The number of mitoses per 1000 cells in the lower layer is thercfore ta7o to three times as great as that per 1000 total cells. This is in accordance with the fact that the number of total cells is about three times as large as the number of cells in the lou7er layer and that the mitoses found in the epidermis above the lower layer are relatively few. THE ANATOMICAI, XBCOKI), VOL. 4'2, NO. :% 228 LEO LORR A N D FILANC’ES 1,. HAVEN S c a s o w l cell p w w t h in s k i n . 111 order to dctermine the cflt’ects of seasoiial changes on the proliferative activity arid striicture of the skin, we diviclcd tlie year into three periods: 1) ?lie summer moiiths, cJune, ,July, August, aiitl Sq)teml)cr; 2 ) tlie wiiiter months, December, January, February, aiid March, aiid, 3) the iiitermetliatc months, comprising April, Alay, O c t o h r , and Novemher. Fifty-five normal animals w e ~ eexamined (table l),two during the summer months, twiity-oiw thiriiig the winter. months, and thirty-two during t lie iiit e rmedi at e moiit h s. a. Seasonal variatioas in the ~iuml)erof cells. The averages foiiiid were as f o l l o :~ ~During summer months : total number of cells per 1 mm., 45’7 ; c:ells per 1 mm. in the lower Iayei-, lW, tiiid, in tlie upper layers, 2Yi. l h r i i i g tlic winter moiitlis: total cells per 1 mm. skin, 484; cells in the lower layer, 159, and cells in the npper layers, 325. l>uring tlic iiitcrmecliate moiitlis: total cells per 1 mm. skin, 460; cells in tlic lower layer, 1.38, aiid in tlic upper layer, 302. The number of cells in the lower layer is approximately the same in all three groups, but tlic number of cells in the upper layer is ,qi.c~~test in the wintc.r months mid lowest in tlie s;ummer moiitlis. This similarity in the iiiimber of cells in the low-cr layer during tlie various seasons of the year and the greater iiumher of cells in tlie upper rows in the winter and iiitermetliate months, as compared with tlie summer months, is prohahly iiot accidental, I~ecanscit is fouiitl equally i l l all the other groups of guinea-pigs whicli we examined from this point of view. It means that dnriiig the summer months tlie cells kcratinize relatively more rapidly than iii the cooler months of the year. If the proliferative activity of the skin is less iii summer time, tlic process of kcratiriizatioii may require about the same time, following removal of the cells into the upper layers, as during tlie rcst of the year, h i t the pushing of the necessary n u m l m of cells into tlie upper rows takes place less rapidly. On the other hand, if tlie proliferation is equally great iii the summer and in tlic winter months, then the process of keratiiiizution proceeds in the individual GROWTH OF THE MPLDEKMIK 229 cells witli greater rapidity iii tlie warmer season. A study of the mitotic activity during tlie different periods of the year should decide between these two possibilities. 1). Seasonal variations in the number of mitoses. I h r i i i g the summer months: Mitoses per 1000 cells, 2.92; per 1000 cells of the lower layer, 5.39. Mitoses per 1 mm. skin, 1.34; in the lower layer per 1 mm. skin, 0.86, and in the upper layer, 0.48. During the winter months : Mitoses per 1000 cells, 5.59; per 1000 cells of the lower layer, 13.91. Mitoses per 1mm. skin, 2.i3 ; in the lower layer per 1mm. skin, 2.18, and in the upper layer, 0.55. During the intermediate months : Mitoses per 1000 cells, 4.80; per 1000 cells of the lower layer, 11.23. Mitoses per 1mm. skin, 3.25; in the lower layer per 1mm. skin, 1.77, and in the upper layer, 0.48. Mitotic proliferation thus is most marked during the winter months, but it is almost a s high in the intermediate period. It is about one-half as great during tlie summer mouths a i d less than one-half, if we consider merely the lower layer. Tlie differeiice between counts per 1000 cells total and per 1000 cells in the lower layer, during summer months and during the rest of the year, is due to the smaller number of cells in the upper layers during the summer. The figures for mitoses per 1mm. skin, instead of per 1000 cells, a r e also highest during the winter and lowest during the summer months. Wliile in this case, as well as in the majority of other groups which we studied, the proliferative activity of the skin seems to be less during tlie summer months, this does not seem to hold good in the case of all the individual groups of animals examined by us. Variatims iw, the epidermis in diferrnt years. Above we have reported on Ihe tests which we made concerning the reliability of our methods by making two counts and by comparing the figures obtained when both ears of the same guinea-pig were examined. There is still another method by means of which we can test the degree of trustworthiness of our counts, namely, by determining the differeiice between 230 LEO LOEB AX11 FllANCES L. HAVEN the counts obtained during the first and second years. We made such a comparison in the case of the adult male guineapig. The averages of forty-one male adult guinea-pigs examined during the first year were as follows: Total number of cells per 1 mm., 467; number of cells in lower layer, 157, find in uppcr layer, 310. Number of mitoses per 1 mm. skin, 2.47, and number of niitoses per 1000 cells, 5.17. The average of fourteen normal adult guinea-pigs, counted during the second year, were: Total number of cells per 1 mm. skin, 474; number of cells in lower layer, 161, and in upper layer, 313. Number of mitoses per 1 mm. skin, 2.21, and per 1000 cells, 4.69. There is only a slight difference between the results obtained in the first and second years. The averages for the number of mitoses are slightly lower in the second year, because there are included here the two animals studied during the summer months, which had relatively low counts. 1’. The proliferative activity of the s k i s of yoiirtg guineapigs. We compared with the behavior of the ear skin of adult male guinea-pigs that of the ear skin of very young guineapigs, males and females, varying in weight between 63 and 205 grams and in age from one to thirty days (table 1, lines s, f , u,v). We studied altogether twenty-seven such animals. ‘I’he average number of cells was 410; in the lower layer it was 158, and in the upper layer, 252. Whereas the figure for the number of cells in the lower layer is about the same as i i i the older animals, the nnmbcr of cells in the upper layers is decidedly lower. The proliferative activity in these young guinea-pigs is likewise lower than in the adult, and if we assume that keratinization occurs here a t allout the same timc after. rcmoval of the C C ~ from S the lower layer as in tlic adult guinea-pig, a lower number of upper cells must result. in twenty-two of these animals a classification was made aceording to the findings during the different seasons of tho year. The average cell coiints of these twenty-two animals, irrespective of seasonal variations, were a s follows : 418 total (*ells,158 cells ill the lower, and 260 in the upper, layer. GROWTH O F T H E EPI1)EBMIS 231 I n this more restricted number of guinea-pigs the decrease in the number of cells in the upper layer is quite evident. Eight of these guinea-pigs were examined during the summer months, with an average total number of cells of 398; the number in the lower layer was 161, and in the upper layer, 237. I n the winter and in the intermediate months the average counts of fourteen animals were: Total cells, 431; cells in the lower layer, 157, and in the upper layer, 274. Again we find in the summer months the average number of cells in the upper layers to be less than in the rest of the year. This is, as we shall see shortly, associated in this case with a somewhat more active cell proliferation during the summer months. The decrease in the number of cells during the summer months must therefore be caused by the greater rapidity in keratinization which evidently takes place during this time of the year. If we compare the behavior of the sexes in these twentyseven guinea-pigs, we find the figures in ten males as follows : Total cells, 386; cells in the lower layer, 156, and cells in the upper layers, 230. I n seventeen females the counts were as follows: Total number of cells, 425; cells in the lower layer, 160, and cells in the upper layers, 265. I n the females the number of cells in the upper layer is somewhat higher than in the males-which corresponds to a somewhat greater mitotic activity in the females. While it is at present still doubtful whether the higher figures in thk females have actually this significance, we may at least conclude that in the young guinea-pigs the activity is certainly not greater in the males than in the females. The average number of mitoses per 1 mm. skin in the twenty-seven young guinea-pigs was 1.60 per 1 mm. skin and 3.94 per 1000 cells. These figures are definitely lower than in the adult guinea-pigs-a result which is somewhat unexpected, inasmuch as it is usually assumed that in very young animals the growth of tissues is more rapid than in older ones. We must therefore conclude that a greater intensity of growth is not found equally in all the tissues of young guinea- 233 LEO 1 A O E l 3 A S D FIIANCES L. HAVEN pigs. This tlifierencv in the growth rate between thc immature young and the adult is probably not of ail accidental iititiire, lwcause it is ol-rscrvecl in various groul)s of tlic ~ o i n i g giiiiiea-pigs. The difference hetween tlie adult and yoii~ig gxinm-pigs is greater if we compare the figures for. the mitoses per 1 mm. than if we compare the figiires for mitoses per I000 c ~ l l s . In the former tlie figure is 1.43 times greater antl in the latter 1.22 times greater in the adults than in the young. The iiumber of upper eells, in which mitoses are iiifrequcnt, is greatiir in the adult animals-wliich tends i*el;itively to lower the iiumber of mitoses per 1000 eclls in t h e aclults. Tf we consider merely the group of twenty-two guinea-pigs, we find the number of mitoses per 1 mm. to he 1.T1 ant1 the number. per 1000 cells, 4.14, while the number of mitoses iii the lower layer per 1000 cells in this la!-cr was 8.00. The corrcspoiicling figures iu the tdults are : per 1 mm. skill, 2.40; per 1000 cells, 5.04, and per 1000 cells of the lower row, 12.04. Again we find the difference between the figures for mitoses per 1000 cells, in young and adult, smaller than the corresponding figures per 1 mm. and per 1000 cells of t h e lower row, because in tlie adults the total riumher of cells m d , especiaII:-, the nnmber of cells in the upper layers are greater tliaii in the young guinea-pigs. The proportion between the number of mitoses in tlic young antl in thc. adult is approximately 2 : 3. If we compare the number of mitoses (liiriiig the summer months a n d dnring tlic rest of thc year, we find f o r the summer months 2.04 mitoses per 1 mm. skiii a t i d 5.05 per 1000 cclls, w h i l ~for. thc rest of the year they are 1.51 a i d 3.63. The figures per 1000 cells accentnate the tliffcrciiccs found, 1 ) e c . a the ~ ~ ~numbel* ~ of cells in the summer months is lower than during the rest of the year. The proportioil hetween tlic rnitosw cluring tlic summer and tlic rest of the year is therefore approximately 4 : 3. It is possible that in \'cry youiig animals a liiglier temperttture is more favorable t l i a i i in adults, but, considering the relatively small figures with wliicli we have to deal in this case and the not very great differences wliicli we found, we must leave at present the question open as t o the reasonal efiects oil cell proliferation in very young guinea-pigs. If we compare the figures for males a i d females, we find iii ten males 1.11 mitoses per 1 mm. and 2.90 mitoses per 1000 cells; in seveiitcen females, 1.89 per 1 mm. and 4.55 per 1000 cells. Again we may coiicliicle that the niimher of mitoses iii the yoiuig fcmalcs is a t least not lower than in tlie young males. To summai*ize,we may state that iii young guinea-pigs tlie iiiimher of cells in the lower layer of tlic epidermis is about the same a s in the adults, hiit that the number of cells in tlic upper lay,-cr is decidedly lower in the yoixiig animals. Furtliermorc, that tlic mitotic activitp is lower in the young than in the mature guinea-pigs, antl the iiiim1)er of (+ellsis smaller iii the summer moiitlis than in the rest of tlie y e a r ; we may also conclude that tlie iinmher of mitoses is not smallcr in the youiig females than in thc. yoiiiig males. Rut ~ v l i e t l i ~itr is actually larger in tlici yoiiiig females tliari iii tlic young males and whether in the 7oiiiig guinea-pigs it is actiially greater i n tlie summer moiitlis tlinii during the rest of tlie year will need furtlier investigation. TYe liave wen that within tlie first weeks of life the strncturc aiitl p~olifei*ativeactivity of tlie epidermis of tlie ear of the guiiie;i-pig dificr from those in adult guinea-pigs, if we tlefine the latter as iiiclntliiig all those animals which a r e sexually mature o r perliaps also those which a r e entering the period of sexual maturity. In early life the proliferative activity aiid the iiumher of upper cells iii tlie epidermis are diminislied. In stiidying the effect of weight in adult guinea-pigs, if we divide tliese animals into two classes, one comprising those below 300 grams in weight and the others ahove this weight, tlie former class shows a greater proliferative activity, but the difference hetwecii the riumher of cells is very s1i ght . s. 234 1,EO LOER ANL) FRANCES 14. HAVEN However, in indiviclual members of each group there is a great variation in the mitotic activity, ranging among lighter (younger) animals between 0.38 and 8.37 and among the heavier (older) animals between 0.38 and 5.32 mitoses per 1 mm. skin. I f , instead of tlie groups a s a whole, we consider the individual guinea-pigs, there is no direct relation between weight a i d proliferative activity of the epidermis. By raising tlie dividing point between the two classes from 400 to 450 grams, the diff'erence in the mitotic activity per 1 mm. skin tliminislies somewhat, but it is still present. The number of mitoses per 1 mm. ekiii in twenty-four animals below the weight of 450 grams is 2.87, and in fifteen animals above the weight of TABLE 2 NChCRER OF GCINEA-l'I(iS Relow $00 gr:lIKlS (average wcigiit , 382 grams) I! . o~~~~~ ' I Above 400 grams 470 41 46.5 .... ' NUYHGR O F CELLS IN LOWER LAYER 1 156 I 159 NUMHKR O F CELLS I N UPPER MIToSEs LAYER "IJb. "' ~ I ' 314 I 308 I I 3.34 1.77 450 grams it is 1.83. It appears again that weight (age) of the gninea-pigs is at least one of the factors which iiifhence Il.. rile - _A!-.. -2 A L - ---Z.J Ip r u i i i_e_m i i u ~ i01 i ~ i eepiuenuis ..---l?l? Z- 111 --:-,.1-. iiiese aiiimais ; LI I- U- LL A L :A !,. L L i~ not the only factor which is of importance. The conclusion that weight (age) of the guinea-pig is one of the factors in this process is confirmed if we further subdivide the normal male guinea-pigs according to weight and indirectly according to age. This table indicates that in mature animals the proliferative activity of the epidermis decreases with increasing weight and age and that the range of variability likewise decreases a s a result of a diminution in the maximum of proliferative activity that is reached in individual guincapigs. The average proliferative activity is lowest in animals ahove the weight of GOO grams, and tlie loss in average pro- 235 GROWTH OF T H E EPI1)EI:MJY liferative activity seems approximately proportional to the increase in average weight. It will, therefore, henceforth be necessary, in testing the effect of various internal and external factors on the proliferation of the epidermis, to consider the weight and age of the guinea-pigs which are used, in order to avoid erroneous conclusions. While there is thus a relation between weight (and age) and cell proliferation in the epidermis in general, there does not need to be correspondence in individual cases, where we find great variations. - - -. - I. Below the weight of 400 grams 3.34 0.38-8.37 Ia. Below the weight of 450, I 24 2.87 0.38-8.37 11. Between the weight of 450 and 550 grams i Ira. Between the weight of 16 2.07 0.64-5.32 2.07 0.64-5.32 1.61 0.38-3.65 grams 450 and 600 grams 111. Above the weight of 550 grams IIIa. Above the weight of 600 1 22 I 17 1 This relationship seems, therefore, to hold only if we consider larger averages. I n a general way it is a well-known fact that mitotic activity of various tissues decreases with increasing age, and only recently Thuringer(7) adds to this evidence; he finds a much greater number of mitoses in the human prepuce of a seventeen-day-old, than in a three-yearold, child. After we had determined that, with increasing weight and age, the proliferative activity of the epidermis of adult male guinea-pigs diminishes, it was of interest to determine whether, parallel with this decrease in proliferative activity, the number of cells in the upper lagers also diminishes. 236 LEO 1,OER A N D IViANCES L. HAVEN We may conclude from this table that, i i i contrast to the proliferative activity of the epidermis, the number of cells in the upper layers remains about the same in the different age classes, a t least until we reach the weight of 600 grams. I11 the guinea-pigs weighing above 600 grams a relatively slight ck~creasein the. numbor of (*ellsdoes possibly take p l a c ~ . 'rhcre is a vcry slight decrcase rioticeahle also in the groiip of animals weighing above 550 grams; hut this is clue merely to the fact that this class includes the guinea-pigs weighing ahove 600 grams. If, then, the numher of' cells in tlie epiclermis remains approximately coiistant with iiici.easing age, then we must assumc? that the result of the increased proliferation ill the groups of younger animals is neutralized by a more rapid keratinization of the upper cells, whereas with a dccrcasing proliferative activity there is associated a slowiiig up of keratinization, until at last, perliaps, the diminution in the proliferation becomes so grcat t h a t the tlimiriiitiori iii the rapidity of kcratinization can no longer eiitirelp compensate for it, and thus a slight, falling off in the number of cells occurs. However, even if the slight decrease in tlie niimber of cells is real, the conclusions as to the re1a t'1011 betweeii proliferative activity a i d rapidity of keratinization can at present be considcretl merely as tentative. of low t c m p w a t i i w 012 the prolifcJratice actirity of the epit?m-m,i.c. In a preliminary experiment seven guinea-pigs, varying in weight hctwccn 325 aiid 460 grams, were kept for a period of six to nine hours in an artificially cooled room at a temperature of ahout 62°F. I n one animal we removed a piece of ear skin before the transfer of the guinea-pig to a warmer room; in the other cases the pieces of skin were talien at once after this transfer had been made. In the first animal the cell counts were rather high, the niimher of cells peT 1 mm. being 502; the number of cells in the lower layer, 173, and in the upper layer, 329. The mitotic activity was low. The figure for mitoses per 1 mm. was 0.32; for mitoses per 1000 cells total, 0.63, and per 1000 cells of the lower layer, 1.84. These figures are decidedly below any figures obtained in the case of other normal guinea-pigs. They indicate, a s might be expected, that if the temperature of the ear is sufficiently lowered, the mitotic activity decreases aiid perhaps also the rapidity of the transformation of skiii into keratin. However, if the temperature of the skin is not sufficiently lowered, or if the skin is allowed t o warm up again after return to a warmer room, the proliferative activity is found to be normal. Thus the averages of the seven guineapigs used in this experiment were found to be approximately iiormal after a previous exposure in the ice box. In this group the average for the number of cells was 484; of the cells in the lower layer, 164, and in the upper layer, 320. The number of mitoses per 1 mm. skin was 2.40, and per 1000 cells, 4.83. SUMMARY A N D CONCLUSIONS 1. We determined the range of error inherent in our method of estimating the proliferative activity of the epidermis aiid the number of cells which are produced before the process of keratinization sets in. F o r this purpose we made two counts in one area of the skin and we also compared the conditions found in both ears. The absolute differences 238 LEO LOEB AND FltANCES L. H A V EN in mitoses here observed are very small; also, the percentage differences are relatively small a s compared with differences found in guinea-pigs which were unlike in certain respects, as, for instaiicc, sex and age. The number of cells in the lower layer of the epidermis is, in the same species of animals arid at a corresponding point of the body, almost a constant, whereas the number of cells in the upper layers differs to a somewhat larger extent. I n the case of the r a t the differences found in comparing both ears were only slightly greater thnn those observed in the guinea-pig. 2. The skin from different parts of tlie guinea-pig differs in proliferative activity and in structure. The chest skin of this species behaves in a similar manner to the skin of ear and chest in the rat and to the chest skin of birds. The proliferative activity and the number of cells pushed into the upper layers, in tho latter types of skin, are much smaller than in the ear skin of the guinea-pig. However, some minor differences exist also in the second type of epidermis. We may assume that two factors determine the numhcr of cells which may be found in the upper layers, namely, 1)the proliferative energy of the epidermal cells, and, 2 ) a tendency to undergo keratinization which is specific in different species. 3. Wc find, then, among tlie animals examined, two types of epidermis, one represented by the epidermis of the ear of the adult guinea-pig and the other by that of the chest of the guinea-pig, rat, and birds. There is a correspondence between the increased proliferative activity and the larger numher of cells we find in the former a s compared with the second group, where number of cells and number of mitoses, as well, are diminished. The decrease in the numbcr of cells occurs hoth in the lower layer, which is mainly concerned in the proliferation, and in the upper layer. However, the greatest difference is observed in the upper layers. This is presumably due to the fact that there exists much less difficulty in increasing the number of cells in the direction from below upward than in the sidewise direction, where the pres- GROWTH O F THE EPIDERMIS 239 sure which the cells exert upon each other increases considerably with even a relatively small increase in the number of cells. This rapidly increasing pressure in the lateral direction may also be at least partly responsible for the polarization of the cells, under normal conditions, the mitotic spindles being oriented in such a m7ay that the proliferation takes place in an upward direction. However, it is probable that this polarization depends mainly on the difference in the conditions which exists between the upper and lower sides of the epidermal cells; the former is directed toward the side from which the oxygen and the foodstuff8 dissolved in fluid are carried to the cells, while the latter is directed away from this source and toward the air surrounding the organism. Polarization is also indicated by the distribution of the pigment which is situated around the outer side of the nucleus. However, this polarization changes as soon as the cells begin to carry out movements in a lateral direction, as in the case of wound healing. The differences which we found between different types of epidermis are of interest in still anotlier direction. There are two w a y in which we may consider the constant cell proliferation going on in the epidermis : 1) It may be thought of as a regenerative process, caused by the constant loss of cells in the upper layers, owing to the keratinization which occurs here; and, 2) it may be considered as a primary condition, dependent on factors in the inherited constitution of the organism, which cause primarily a cell proliferation and which secondarily place the cells thus pushed upward in such a c,onclition that they must undergo degenerative processes in the form of keratinization. Our investigations show that the second alternative is the essential factor underlying this process, and this conception is confirmed by other facts which show that the stimulation to mitotic prolifcration is primary and the keratinization is secondary. We may mention in this connection the cyclic processes taking place in the vagina of the guinea-pig a i d the effect of mechanical irritation on cyliiiclrical epithelium which map transform it into squamons(~1.11e pi t 11eli um. 4. There occur several variations in the cordition of the epidermis during the summer months. The number of cells in the upper layers is smaller then than during the rwt of tlw pear. This diffcwmx is found throughout all of our groups. There is in general, also, a tendency toward a diminishccl proliferative activity of the cells during the summer months ; this latter condition, however, seems not in every case to be associated with a lowering of the number of cells. ‘l’lic lowered proliferative activity can, at least in part, acc w u i i t foi. tlie lower number of cclls in tlie uppcr laycm, altliougli in some cmes perhaps tin increased tendency to 1it.i.ntiiiization may he responsible for it. It is probaI)le that tlie very marked rise in temperature, which occurs during the summer months and which leads to a lowering of all tlie activities of the guinea-pig, is mainly responsible for this c+liange. 5. While in very young, sexually immature guinea-pigs the iiumber of cells in the lower layer of the epidermis is about tlic same as in the adults, the number of cells in the upper lilyers is defiiii tely lower. The mitotic activity is likewise lower in the very yoling, and to it corresponds the rclativcly small number of cells in the upper layers. Rizcli a result is iiiiexpwtcd, in view of the fact thal tlic geiieral growth of ilicse poiiiig animals is very i*apitl,and we must assume that some special factors are responsible for this condition. It may be suggested that the same factor may liliewise hc caoiiwrned iii tlie iiicreasetl proliferative activity in these animals fouiid tluring’tlic summer months. However, the comparative prolifei-atioii in summer and winter months in these guineapigs iieeclcj further investigation. In coiitradistinction to conditions which we shall later dcscribe in the adult guinea-pig, the proliferative activity is not greater in tlic males than in the females. 6. There is a gradual decrease in the proliferative activity of tlie epidermis of the ear in the adult guinea-pigs with GROWTH OF’ THE EPIDERMIS 241 illcreasing weight aiid age. The curve representing the proliferative activity of all guinea-pigs, ill regard to weight and age, shows, therefore, an ascending branch, representing the conditions obtaining in animals within the first few weeks of life, and a descending branch, representing the conditions found iii older animals. I n aiialyziiig the eflfect of factors influencing the proliferative activity of these animals, it will, therefore, be necessary to take the factors of weight and age into consideration. The numher of cells in the epidermis represents a more constant coriciition than the proliferative activity. I n order to explain this fact, we may tentatively assume that an iiicrease in tho proliferative activity of the younger individuals, within the same group of guineapigs, is associated with a greater tciidericy to uriciergo kera tinization. LITEEATURE CITED 1 LOEB,LEO,AND ADDISOS, W. 11. E’. 1911 Arch. f. Eiitwicklgsrncrli., Hd. 32, 8. 44. 2 LOER,LEO 1920 Journ. Metlir:rl Research, 1-01. 41, p. 247. 3 ADDISON,W. H . F., AND LOEH,LICO 1913 Arch. f . Entwieklgsmcc.l~, Rd. 37, S. 633. 4 SPAIN,K. <’. 1915 Journ. Exper. Medicine, vol. 21, p. 193. 5 RPAIK, K. C., A N D LORI(,LEO 1918 Jourii. Exper. hlcdiciiie, vol. 23, p. 107. 6 AKAIWA, H-4cHrao 1919 Journ. Mrdieal Itesc:ircli, vol. 40, pp. 311, 371. 7 THURINCIER, 6. M. 1924 Anat. Bee., vol. 28, 11. 31.