The Postnatal Growth of the R a t Lung' I. MORPHOMETRY PETER H. BURRI, JAROSLAV DBALY AND EWALD R. WEIBEL Department of Anatomy, University o f Berne, Berne, Switzerland ABSTRACT The postnatal growth of the lung was quantitatively investigated in rats aged 1 , 4 , 7, 10, 13, 21,44 and 131 days by light and electron microscopic morphometry . Lung volume (V,) increased first directly with body weight (W). After day 10 V, followed the function WO.?O.Based on the quantitative findings the postnatal lung growth could be divided into three phases. 1. Lung expansion (up to day 4 ) : Lung volume increase resulted almost exclusively from an 87% enlargement of the existing air spaces. 2. Tissue proliferation (day 4 to 13): All tissue compartments showed a pronounced mass increase, followed by a high gain in capillary volume. Alveolar and capillary surface areas (Sa, Sc) developed rapidly due to subdivision of the primitive air sacs. 3. Equilibrated growth (third week to adult age): An initial period of redistribution of tissue mass with septa1 lengthening and further rapid increase in Sa and Sc was followed by proportionate alveolar growth. In the adult further lengthening of the interalveolar septa or continued alveolar formation could not be excluded. During the period of fundamental internal remodelling of the lung, its function, as determined by the morphometric pulmonary diffusing capacity, was not impaired. Until recently the question of whether the development of the mammalian lung ceased a t birth was unsettled. Based on adult to newborn alveolar size relationships, Koelliker (1879) stated that the structure of the adult lung resulted from simple enlargement of the structures existing at birth. His concept, supported by the work of others (Willson, '22; Barnard and Day, '37; Norris, Kochenderfer and Tyson '41) was opposed to the view that new respiratory units are formed during postnatal growth of the lung (Broman '23; Heiss '23; Bremer, '35, '36/'37; Willson '28; Cohn, '40; Engel, '47, '53; Clemens, '55). Today this problem seems to be solved. Within the last fifteen years, more and more evidence indicates that in most mammalian species new alveoli are formed postnatally and hence the newborn lung cannot be considered to be a miniature of the adult one (Dingler, '58; Emery and Mithal, '60; Boyden and Tompsett, '61, '65; DunANAT. REC., 178: 711-730. nill, '62; Neuhauser, '62; Emery and Wilcock, '66; Reid, '67; Weibel, '67a; Davies and Reid, '70; Burri and Weibel, '71a). Disagreement is still great, however, on the extent and mode of this postnatal structural rearrangement and on the age at which this process is completed. Hoping to elucidate some of the events involved in postnatal lung growth we attempted a systematic analysis of the rat lung during the first three weeks after birth. Our approach was based on three methodological facets : ( 1) a quantitative investigation by means of stereologic techniques; ( 2 ) a n autoradiographic analysis; ( 3 ) a light and electron microscopic purely morphological study. This paper presents the quantitative data obtained on the growing rat lung. It should provide the basis for the understanding of the autoradiographic findings Received July 31, '73. Accepted Oct. 12,'73. 1This study was supported by grants 3.5.68 and 3.682.71 from the Swiss National Science Foundation. 711 712 P. H. BURRI, J . DBALY AND E. R. WEIBEL as well as of the morphological alterations described in two further publications (Kauffman, Burri and Weibel, in prep.; Burri, in prep.). a t 180 EL intervals in a plane parallel to the inferior surface of the lobe. They were sequentially arranged on a glass slide and stained with hematoxylin-eosin. MATERIALS AND METHODS D. Specimen preparation for electron microscopic investigation A. Animals Young rats from about seven litters of the CFN-COBS-strain were killed in groups of three at the age of 1, 4, 7, 10, 13 and 21 days, day 1 corresponding to the age of 24 -t 6 hours. No distinction regarding the sex of the animals was made, since there are no differences in growth rate between sexes in the early postnatal period (Grab and Munker, '64). To facilitate interpretations, data from rats, 44 and 131 days old, have been included in some tables or in the text. They were obtained in previous studies with identical techniques on rats of the same strain but i n another context (Burri and Weibel, '71a,b). The group of day 131 exclusively comprised male animals. B. Lung fixation The animals were anesthetized with Hypnorm" (FentanylO.2 m g and Fluanison 10 m g l m l ) in a dose of 0.2 mljlOO gm body weight, diluted 1:lO with saline. After tracheotomy and careful perforation of the diaphragm to create a pneumothorax, their lungs were fixed by intratracheal instillation of glutaraldehyde at a standardized pressure of 20 cm of water in a supine position, according to techniques used previously (Weibel, '70/'71; Burri and Weibel, '71a). When flow of fixative stopped, the trachea was ligated, the chest organs were removed and stored in toto for two to four hours in the fixative at a temperature of 4°C. Then, the lungs were neatly dissected and their volume measured by saline displacement (Scherle, '70 ) . C . Specimen preparation f o r light microscopic investigation A light microscopic morphometric investigation of the lung was performed, using step sections through a whole lobe. For this purpose the right middle lobe of each lung was embedded in paraffin wax; step sections of 4 thickness were obtained For electron microscopic morphometry the lower part of the left lung was taken as a representative sample for the analysis of lung parenchyma. The reasons for this choice have been discussed previously (Burri and Weibel, '71a). Slices of the lowerehalf of the left lung were cut into small blocks, washed in phosphate buffer for one and one-half hours and postfked in O,O, for two hours. They were given a second wash in maleate buffer, then immersed in 0.5% uranylacetate (block contrast) for one hour at room temperature, dehydrated in a graded ethanol series, and embedded i n Epon 812 (Luft, '61). Large sections of regular thickness, cut on an LKB-Ultrotome 111 with diamond knives, were picked up on Parlodion 200 mesh grids and additionally stained with leadcitrate (Reynolds, '63). Blocks containing throughout large vessels or bronchioles were discarded according to the morpliometric standards established below. E. Solutions The osmotic pressure of the following solutions was adjusted to 340 mosm except for the uranyl-acetate block-contrast step. 1. Glutaraldehyde fixation For lung fixation a 2.5% glutaraldehyde solution in 0.03 M K-phosphate buffer at pH 7.4 was used. Prior to instillation the solution was cooled to 4°C. Postosmification The washing of the blocks between aldehyde and osmium fixation was performed in a n ice-cooled 0.2 M K-phosphate buffer of pH 7.4. The blocks were postfixed in a chilled 1% OsOl solution in 0.2 M s-collidinejHC1 buffer a t pH of 7.4. 2. 3. Block staining The washing between osmium fixation 713 MORPHOMETRY OF POSTNATAL LUNG GROWTH and block-staining was done in 0.05 M maleate-NaOH-buffer of pH 5.0. For blockstaining 0.5% uranylacetate was added to the above maleate buffer. cytes); ( 2 ) the alveolar and capillary surface areas; (3) the harmonic mean thickness of the air-blood barrier (Weibel and Knight, '64). F. Morphometric analysis ( a ) Light microscopic evaluation The major purpose of these measurements was to establish, whether there was a shift in the volume density of lung parenchyma, i.e., the gas exchanging parts of the lung, during growth. Step sections were evaluated at a magnification of 200 times, using the Wild M501 automatic sampling stage microscope (Weibel,$70). All sections were systematically covered with a square test point lattice with a period of 550 p. This was performed by marking just one test point in the centre of the test-screen and by traversing the section automatically and stepwise by one width of the test-screen about every 1.5 sec. The volume densities of the following compartments were determined : lung parenchyma (= gas exchange region) and non-parenchyma. The latter comprises the connective tissue septa, the airways down to the respiratory bronchioles, and the blood vessels down to a diameter of 25 I". Furthermore the point hits on air and tissue in the whole lung and on lumina of the blood vessels and of the airways in the nonparenchyma were recorded. (c) ( b ) Electron microscopic evaluation In most cases three blocks from each lung were randomly chosen to be cut and from every block one section was taken to be photographed. From each section 12 pictures were taken in the mesh-corners of the supporting grid on a Philips EM 200 at a primary magnification of about 1200 times. The photographs were analysed at a final magnification of 11,000 times on a projection screen containing a multipurpose test system with 84 lines and 168 test points for point and intersection counting (Weibel, '69). The parameters estimated were : ( 1) volume densities of the air spaces, of tissue and its components (epithelium, interstitium and endothelium), and of capillaries and their content (plasma and erythro- Calculation of pulmonary diffusing capacity We calculated for each animal the pulmonary diffusing capacity (DL) using the following equations (Weibel, '70) : (1) 1 1 1 1 DL Dt Dp De -=-+-+- where the right hand components, the diffusing capacities for tissue (Dt) plasma (Dp) and erythrocytes (De), were obtained by : 1 ( 2 ) Dt == Kt . 1/2(Sa + S c ) . Tht (4) D , = O . V , The structural parameters (alveolar and capillary surface areas S, and S,, the capillary volume V, and the harmonic mean tissue thickness T h t ) were morphometrically determined for each animal. For the harmonic mean plasma thickness T h p a constant value of 0.18 was assumed. This value represents the mean obtained from measurement on a limited number of rats. The choice of the physical coefficients K,, K, and 0 influences the absolute values for the diffusing capacity; following our practice for comparative studies, the highest values found in the literature were applied in this study (Weibel, '70/'71, Burri and Weibel, '71a, Siegwart, Gehr, Gil and Weibel, '71). @ (d ) Data processing The point hits and intersection counts were registered by means of a n electrical counter with automatic data transfer to tables and punch cards (Weibel, '67b). Calculations and statistical analysis were performed on a Bull-gamma 30 computer with the Pocoster-program of Gnagi, B u d and Weibel, '70. As a rule a result was considered to differ significantly from another one, if the significance level a of 0.05 was reached or surpassed in Student's t-test. 714 P. H. BURRI, J. DBALY AND E. R. WEIBEL TABLE 1 N u m b e r o f animals, body weights, lung volumes and specific lung volumes for the various age-groups. Values are m e a n s 2 1 standard error. Significant differences between groups are given f o r specific lung volumes Group Ane in davs Number of animals 1 3 4 3 7 3 10 3 13 3 21 Body weight Lune volume Specific lung volume gm cm3 cm3/lOO gm * 7.2 0.1 12.0 2 0.2 15.6 2 0.2 18.5 & 0.3 0.57 & 0.01 0.90 7.89 t 0.32 - 7.49 No * 0.03 2 0.42 1.22 t- 0.05 2 0.30 1.44 t- 0.07 & 0.29 7.82 25.7 1.80 & 0.06 7.00 &0.10 3 38.5 2 0.1 2.25 -to.11 44 8 140.5 t 7.1 131 3 456.7 12.0 RESULTS A. Body weight and lung volumes The data for the body weights ( W ) and lung volumes (V,) as well as the specific lung volumes (i.e., lung volume per 100 gm body weight) are contained in table 1. During the first ten days the increase of lung volume was directly proportional to the increase in body weight, a s is reflected by the constant values of about 7.8 cm3/100 gm for the specific lung volume (V,,/W) and by a n exponent of about 1 in the allometric function (fig. 1). After day 10 the lung volume increased at a less steep rate (fig. l ) , the values for VL/W falling gradually to 5.86 cm3/100 gm on day 21 and to 2.93 cm3/100 gm at the age of 131 days. B. Morphometric results 1. Light microscopic morphoinetric results A synopsis of the compartmental distribution of volume densities in the various age groups is presented in table 2. The parenchymal volume density (V,.,) ranged between 0.82 and 0.85 and showed no definite trend. No 7.80 2 0.7 2 Difference to previous group sienificant? No Yes P < 0.05 r 0.44 P < 0.05 2 0.20 4.54 rO.ll P < 0.005 13.37 t 0.52 5 0.04 P < 0.005 6.34 Yes 5.86 Yes Yes 2.93 The volume density of total tissue in the lung (V>7tl)ranged from 0.29 to 0.38; the relative amount of air in parenchyma and non-parenchyma (V,,,) fluctuated from 0.54 to 0.66 and inversely with VVtl.For both parameters, the only significant change could be located between days 4 and 7, where relative tissue mass increased by 17% at the expenses of the air compartment. The volume density of blood vessel lumina (V,.,,) showed quite large variations during the first three weeks; this could however be due to chance, as indicated by the large standard errors of ?.Three different types of results must be considered: ( a ) Relative results These represent the primary morphometric data reflecting the relative volumes (= volume densities Vv) of the compartments of the lung and the relatlve surface areas ( = surface densities S V ) , i.e., the surface areas per unit lung volume. The relative results a r e directly derived from the point counts (Glagoleff, ,'33; Chalkley, '43) and the lntersectlon counts (Tomkeieff, 451, obtained by superimposing the test screen to the image to be analysed. (b) Absolute sesults Except for the harmonic and arithmetic mean thickness, these results are calculated by multiplication of the relative results with the absolute volume of lung parenchyma (= total lung volume x percentage of lung parenchyma). ( c ) Specific results These express the absolute results with respect to body weight. They are given in volumes (cm3) or in surface areas (m2) per 100 gm body weight. 715 MORPHOMETRY OF POSTNATAL LUNG GROWTH 50 VIL cm 3 20 10 W z 5 3 014 1 , 2 5 10 20 50 100 200 500 . DO BODY WEIGHT Fig. 1 Bilogarithmic plot of lung volume (V,) against body weight ( W ) . Two growth phases can be distinguished (days 1 to 10 and 10 to 1 3 1 ) with two distinct allometric relationships. The points represent the means of the age groups listed in table 1: brackets include ? 1 standard error. The allometric functions were calculated with the values of the individual animals. some groups. On day 131 this parameter was however significantly increased ( P < 0.005). The volume densities of bronchial and bronchiolar lumina (V,,,) decreased significantly from 0.059 at day 1 to 0.042 at day 21, a reduction of more than 28% ( P < 0.05). On day 44 it was only 0.032, and on day 131 i t was once more significantly reduced to 0.023 (P < 0.05). 2. Electron microscopic morphometric results ( a ) Relative results The compartmental volume distribution within the parenchyma of the lungs of the various age groups is given in table 3. The volume density of air spaces (Vva) increased between days 1 and 4 from 0.63 to 0.75 and was reduced again from day 7 on to values around 0.69. On day 44 this parameter was significantly higher by about 15%. The volume density of tissue (Vvt) varied in the opposite sense: between days 1 and 4 there was a significant fall from 0.28 to 0.19, followed by a n increase to 0.25 on day 7. After day 13 i t was gradually reduced and reached 0.087 on day 131. The tissue comprises a n epithelial, an 716 P. H . BURRI, J . DBALY AND E. R. WEIBEL TABLE 2 Light microscopic evaluation of relative quantitative composition of lung tissue. V o l u m e densities (Vs) are given as m e a n s f 1 standard error. p , parenchyma; n, non-parenchyma; t l , tissue in whole lung; la, lung air spaces ( i n c l . a i r w a y s ) ; vl, lwmina of blood vessels ( d o w n to 25 p); bl, l u m i n a of airways ( d o w n to respiratory bronchioles).VVp+ Vs,= 1 . VVti+ V,(, + V l U r= 1 . Significant differences between groups with asterisks: * p 0.05; * * p 0.005 < < ~~~ ~~ ~ Parameters Group Age in days 1 4 V\ u VV” 0.822 2 0.006 0.178 f0.006 0.318 -t 0.015 0.824 0.176 20.013 * 0.003 2 0.013 7 10 13 21 0.820 0.180 -t 0.010 f0.010 0.295 VV”1 VVbl 0.047 t 0.010 0.059 * 2 0.005 ** 0.047 t 0.007 0.055 t0.007 ** 0.061 t 0.009 % 0.002 VVI, VVti 0.635 f0.009 ** 0.658 f0.006 0,357 :‘* % 0.009 0.582 i0.009 0.817 0.183 0.375 0.544 0.081 f 0.026 20.026 2 0.026 ? 0.032 2 0.018 0.049 0.037 t 0.004 0.854 0.146 0.339 0.605 0.056 0.037 f0.011 2 0.011 f0.035 ? 0.036 & 0.002 % 0.003 0.823 0.177 f0.015 0.289 t 0.021 0.647 0.064 0.042 -C 0.025 -t 0.007 f0.004 f 0.015 * TABLE 3 Electron microscopic estimation of volume densities of compartments in lung parenchyma. Volume densities of air spaces ( V l u ) , of capillary blood ( V s c ) , and of tissue ( V s t ) . Values are m e a n s -t 1 standard error. Important changes occur in t h e first week and a f t e r day 21 ( s e e t e x t ) . Asterisks indicate that difference to n e x t younger group i s significant: * P < 0.05; ** P < 0.01; * * * P 0.005 < Group Age in days 1 4 7 10 13 21 44 131 VVa Vve 0.630f 0.020 0.75220.014 * * * 0.687 2 0.027 * 0.691 ? 0.024 0.668 -t 0.024 0.699 f0.044 0.799C0.008 * * * 0.772 f0.005 0.088 f0.002 0.055t0.018 0.064 2 0.012 0.065e 0.003 0.087 f0.008 0.108k 0.020 0.092%0.005 0.141*0.005 *** endothelial and an interstitial compartment. These components underwent a biphasic shift consisting in a relative augmentation of the interstitial volume from day 1 to 7 with a concurrent reduction of epithelial volume density (table 4 ) . The inverse phenomenon took place between days 13 and 21. Both these changes were highly significant for either compartment (P < 0.005). The endothelial volume density showed the same tendency as that observed for the epithelium, however only the increase to day 21 and day 44 reached a significant level. The volume density of capillary blood (VVc,table 3 ) showed at day 4 a fall by vst 0.282 2 0.022 0.19420.005 ** 0.249fO.018 ** 0.244 2 0.025 0.245 t 0.022 0.194 -t 0.029 0.109tO.006 * * 0,0872 0.001 * more than 35% which had recovered on day 13. For a certain time this parameter stabilized around 9-10%. In adult rats however, it was again significantly increased to 14%. and capillary Both the alveolar (S,.) (Sv,) surface densities showed similar relationships with age (fig. 2). The sudden drop by about 30% for Sv. and 40% for S,, between days 1 and 4 was followed by a fast recovery until day 10. On day 21 S\. was enlarged by 47% and Sv. by 30%, when compared to day 1. On day 44 Sv. was significantly reduced to 750 cm5/cm5 (Burri and Weibel, ’71a) and remained at this level until day 131. 717 MORPHOMETRY OF POSTNATAL LUNG GROWTH TABLE 4 Composition of parenchymal tissue. Volume densities of epithelium (V,,,), interstitium (VV,,,) and endothelium (VV,,). Values are means 2 1 standard error. Asterisks indicate that difference to next younger group is significant: * P 0.05; *:'P 0.01; **'<'P 0.005 < < < Group Age in days Vvell vvin Vven 1 4 7 10 13 21 44 131 0.237 2 0.007 0.191 20.009 ** 0.16320.003 * 0.159 rt 0.01 1 0.151 f0.005 0.26320.007 *** 0.274 f0.013 0.282f 0.013 0.578 t- 0.006 0.647 f0.022 * 0.678 f0.017 0.659 0.014 0.664? 0.012 0.514f0.012 *** 0.464 C 0.017 0.463 2 0.018 0.18520.001 0.1622 0.022 0.159& 0.015 0.182&0.004 0.185t0.008 0.223 2 0.012 * 0.261 I?I 0.007 * * 0.255 f0.015 SV m2 4rfi t 0.12 m 2 w 5 W V a LL (L 3 * 0.08 t (L 42 - a Q V n f 0.04 .----a sv, sv C LL 4 s 3 Q 8 I 1 4 I 7 1 1 10 13 I 21 DAYS AGE Fig. 2 Plot of alveolar (S,.,) and capillary surface density (Sv.) against age. T h e rapid fall on day 4 is rapidly recovered. Brackets include 2 1 standard error. (b) , I Absolute results Volumes and surfaces The absolute volumes of the different compartments of lung parenchyma and the absolute alveolar and capillary surface a areas are listed on table 5a and 5b together with the rates of increase. Changes between days 1 and 4. Va increased by 87% from 0.30 cm3 to 0.57 cm' (table 5a j. The other parameters exhibited no significant changes. The increase in v TABLE 5a < < < Total percent change from day 21 to 131 131 Percent change from day 44 to 131 44 Percent change from day 21 to 44 Total percent change from day 7 to 21 21 Percent change from day 13 to 21 13 Percent change from day 10 to 13 10 Percent change from day 7 to 10 7 Percent change from day 4 to 7 1 4 Percent change from day 1 to 4 Group Age in days 66.3 * * * 965.02 35.9 31.0 ** 1277.02 74.2 32.3 *+ 2159.0 C 268.5 69.1 * 0.31720.036 69.5 * 0.365 2 0.036 0.512 C 0.022 40.3 * 0.551 2 0.051 7.6 0.251 2 0.024 73.1 * * 0.289 t 0.034 0.066 2 0.014 _f 0.019 520.1 650.8 159.6 335.2 208.1 *** 96.2 *** 264.9 102.4 *** 7857.5 2 379.3 129.0 *** *** 79.9 3882.9 f 192.2 193.1 15.6 90.0 * * * 1.0472 0.060 2.398 k 0.056 ** 62.7 * * * 59.6 0.565 t 0.040 73.8 0.9192 0.041 143.7 *** 0.480 f 0.022 41 .O -6.1 15.1 1.479k 0.025 *** 198.5 * 0.354 2 0.031 30.4 0.377 k 0.023 15.1 8.130f0.337 216.0 4.1432 0.108 90.0 45.9 0.197 t 0.030 1.31120.173 26.7 77.6 * * * 0.135 2 0.011 15.2 0.076 2 0.005 27.1 * 1.0352 0.068 18.0 * * * 0.814 22.1 * * * 0.690 2 0.013 46.7 736.72 36.8 5.6 7.4 0.0 * 285.2 100.6 *** 8158.72 343.5 92.0 *** 4067.2% 196.9 212.6 59.4 2118.02271.5 48.6 * * * 1329.02 56.2 32.0 894.12 50.2 75.2 * 677.52 96.9 - 9.4 426.4-e 28.1 386.72 105.3 383.22 15.4 442.92 38.9 87.1 * * * _f cm3 0.042 2 0.001 0.042 t 0.014 cm3 0.3022 0.008 0.565 0.013 sc CnY Sa CrnZ VS cm3 0.17720.012 0.187k0.015 Vt cm3 0.13520.012 0.145f0.001 vc Va Absolute volumes and surfaces o f the parenchymal lung compartments. V a , air space volume; V c , capillary blood volume; V t , tissue volume; V s , septa1 volume (volume o f tissue + capillary b l o o d ) ; Sa, alveolar surface area; Sc, capillary surface area. Values are means 2 1 standard error. Significance levels of percent changes: * P 0.05; ** P 0.01; *** P 0.005 719 MORPHOMETRY OF POSTNATAL LUNG GROWTH TABLE 5b Absofute volumes of tissue components. V o l u m e of epithelium (V,,), of interstitium (V,"), o f endothelium (Ve,,). Values are m e a n s e 1 standard error. Significance levels f o r differences: * P 0.05; ** P 0.01; * * * P 0.005 < < Group Age in days 1 4 Percent change from day 1 to 4 7 Percent change from day 4 to 7 10 13 Percent change from day 7 to 13 V,, V., Vi" mm3 3 2 . 0 t 2.0 27.72 1.4 mm3 mm3 78.4& 7.7 94.02 2.6 25.12 2.3 2 3 . 6 2 3.2 - 13.2 + 19.8 - 6.0 40.8% 3.6 $47.2 * 45.5-t 3.9 56.92 3.9 +39.4 '2 21 Percent change from day 13 to 21 93.22 8.5 44 Percent change from day 21 to 44 155.02 12.7 131 Percent change from day 44 to 131 262.4'- 18.9 Total percent change from day 21 to 131 < +63.8 ** 66.3 * 69.3 *** 181.6 lung volume was brought about almost exclusively by the expansion of air spaces, whilst the small augmentation of tissue mass was due to the enlargement of the interstitium (tables 4, 5b). Changes between days 4 and 7. In this period Va increased only by 22% whereas all the other parameters showed augmentations between 47 and 75% (table 5a). The alveolar and capillary surface areas were enlarged by 66 and 75% respectively. Changes between days 7 and 21. This period was characterized by a more equilibrated growth of air and tissue compartments. Before day 13 both showed similar rates of increase: 18% and 15% from day 7 to 10 and 27% and 30% respectively from day 10 to 13. After day 13 the air compartment grew much faster; thus i n the whole period from day 7 to 21 the air space volume almost doubled compared with a n increase of only 41% for the tissue, All tissue components proliferated u p to day 13, the endothelium showed a n especially 170.42 18.2 $81.4 *+ 190.7t25.1 250.5 t 18.1 $47.0 *' 182.52 19.9 - 27.2 : 264.4 224.3 44.9 *' 422.72 2.3 59.9 131.6 *** 4 0 . 1 2 5.0 + 70.2 * 52.3? 5.9 69.32 3.1 $72.9 *** 78.12 3.6 + 12.7 146.02 6.8 89.6 **:' 234.1 223.1 60.3 ak *1 199.7 high volume gain (table 5b). After day 13 only the epithelium increased markedly, whereas the interstitium exhibited a significant loss of absolute volume! It is important to notice the extremely high growth rate of the capillary blood compartment; it was more than doubled between days 7 and 13 and was three times as high on day 21 as on day 7. The alveolar and capillary surface areas increased three-fold between days 7 and 21 (table 5 a ) . The changes which occurred in the alveolar volume and alveolar surface area in the first three postnatal weeks are well reflected by the surface to volume ratio of air spaces (S,/V), (fig. 3 ) ; this ratio estimates the degree of complexity of the outlines of the air compartment. The capillary loading of the alveolar surface area, measured by the ratio Vc/Sa, fell until day 10 and then increased sharply between days 10 and 13. The mean septum thickness, estimated by the ratio of septa1 volume to half the alveolar surface area 720 P. H. BURRI, J . DBALY AND E. R. WEIBEL 0.2-- 2 ( 2 . Vs/Sa), showed a tendency to decrease with a marked fall after day 13. Changes after day 21. Since the increase in lung volume was almost sixfold from day 21 to day 131, all measured parameters were found to be greatly augmented (table 5a). The lowest rate of increase was in the tissue compartment which enlarged by a factor of 2.6, whereas the highest rate was shown by the capillary volume which increased 7.5 times. The alveolar air compartment changed at about the same rate as lung volume, by a factor of 6.2. The alveolar and capillary surface areas were enlarged to a similar degree by factors of 3.6 and 3.9 respectively. Within the tissue compartment Vep and Ven underwent an almost threefold increase whereas the interstitial volume was only augmented by a factor of 2.3 (table 5b). fl Air-blood barrier thickness The arithmetic as well as the harmonic mean barrier thicknesses of tissue de- -0.4 creased progressively (table 6). The arithmetic mean thickness (;) was highest on day 4 with 3.8 p, dropped to 1.7 on day 21 and to 1.15 on day 131. This reduction was reflected by decreases in all three components of the barrier, but the interstitial layer was most markedly reduced as is illustrated in figure 4. The harmonic mean thickness of the tissue barrier TI,^), was lowered from 0.64 on day 1 to 0.39 I.1 on day 21 (fig. 5 ) ; the latter corresponds to values in adult animals (0.40 on day 131, table 6). Y Pulmonary diffusing capacity The absolute values of the morphometric pulmonary diffusing capacity D, is plotted double-logarithmically against body weight in figure 6. The broken line represents the calculated regression line for the whole period investigated (day 1to 131) and follows the allometric equation : D L , - ~ S=~0.0123 . W0.*' where W is given in grams. The correlation MORPHOMETRY OF POSTNATAL LUNG GROWTH 721 - 7q 5.0 z_ m m W z 4.0 Y uI I- - 3.0 Lz Lz Q. m 5 2.0 W I 0 L I 1.0 I c- [L 4 AGE I N DAYS Fig. 4 Arithmetic mean barrier thickness of tissue ( 7 ) and its components epithelium, interstitiurn and endothelium. After day 13 the barrier thickness is greatly and significantly reduced at each age step. Columns are means: brackets include f 1 standard error. coefficient r for this curve was 0.986. A significance test of the exponent showed that the slope of the curve dfiered significantly from 1.0 (P < 0,001). However, analysis of the data from the animals aged day 1 to 21 compared with the data from animals after day 21 showed that there were significant differences in the rate of increase of diffusion capacity. The slope of the curve relating days 1 to 21 was 1.04, whereas the slope after day 21 was 0.73. The difference is significant at P < 0.005. In the first period the slope did not differ significantly from 1.0, whereas i n the later period the difference was highly significant. ( c ) Specific results The results expressed per 100 gm body weight are presented in table 7. The parenchymal portion of the lung increased proportionally to body weight during the first ten days, but later on it lagged behind body growth, resulting i n a lowering of the quotient Vp/W. Up to day 10 the other parameters varied according to the shifts which occurred in their relative values, i.e., either in their volume or surface density. The specific results thus reflect the expansion of the air spaces between days 1 and 4: Va/W increased, but there was a marked diminution in the specific volumes of capillary blood and tissue a s well a s i n the specific alveolar and capillary surface areas. These considerable diminutions however were already recovered to a major part on day 7. If we compare the specific values of day 21 with those of day 44 and of day 131, a drastic and significant fall in all parameters of the older animals can be observed. For some of these parameters (Vp/W, Va/W, Vt/W) this tendency to decrease becomes noticeable before day 21. 722 P. H. BURRI, J. DBALY AND E. R. WEIBEL Th' 08 A z T v, 06 .- Y I! I c c w c 04 4 m Z Q W z 0 z 0 02 I [L 6 I 4 L 7 +/AGE IN DAYS Fig. 5 Harmonic mean thickness of air-blood barrier ( T h t ) . A significant decrease occurs after day 13. On day 21 adult values are already reached. Points represent values of single animals; columns are means; brackets include k 1 standard error. DISCUSSION A. Discussion of previous quantitative findings There are relatively few publications devoted to the quantitative investigation of TABLE 6 Arithmetic (T) and harmonic ( T h t ) mean thicknesses of air-blood barrier. Values are means & 1 standard error. Asterisks indicate that difference to next younger group i s significant: * P 0.05; * * P 0.01; + * + P < 0.005 < < Group Age in days - T P 1 4 7 10 13 21 44 131 3.45'0.29 3.82 ?z 0.75 3.69k 0.21 3.1720.33 2.9620.04 1.6920.14 *** 1.42 2 0.07 * 1.1520.04 Tht P 0.637 i0.050 0.509 f0.045 0.506 k 0.015 0.449 i0.028 0.465i0.013 0.393 2 0.007 * + 0.373 2 0.014 0.40020.012 postnatal lung growth. Dunnill ('62) counted the number of alveoli in lungs of human infants and found that these increased from 24.10" at birth up to the age of eight years, when adult values of around 300 million alveoli were reached. I n parallel he found a ninefold increase in the number of respiratory airways and a linear increase of air-tissue interface with body surface area. Emery and Wilcock ('66) counted alveoli on standardized transsections through the right middle lobe and concluded that in the human lung new alveoli can be formed up to the age of 20. With a somewhat different technique Emery and Mithal ('60) had already described a thousandfold increase in the number of alveoli from birth to puberty. More recently Davies and Reid ('70) confirmed the findings of Dunnill, alveolar number increasing from 17.106 at birth to 336 millions a t the age of 11 years. The first stereologic approaches to post- 723 MORPHOMETRY OF POSTNATAL LUNG GROWTH TABLE 7 Specific volumes and surface areas o f lung parameters during postnatal growth. Specific volume of tung parenchyma ( V p / W ) , of parenchymal air spaces ( V a / W ) , of capillary blood ( V c / W ) , and of parenchymal tissue ( V t / W ) . Specific surface areas o f parenchymal air spaces ( S a / W ) and o f capillaries ( S c / W ) . V / W z cm3/100 gm body weight; S / W = m z / l O O g m body weight. Values are m e a n s k 1 standard error. Asterisks indicate that difference to n e x t younger group is significant: * P 0.05; < ** P < 0.01; *** P < 0.005 Group Age in days 1 4 7 10 13 21 44 131 Vp/W Va/W 6.49 20 . 3 0 6.16 k 0 .3 2 6.42 k0.18 6.35 C 0.04 5.98 * * * kO.01 4.84 k 0.44 3.73 c 0.10 2.30 * * * k 0.03 * *** 4.17 i- 0.1I 4.72 k 0.19 4.42 C 0.05 4.41 20.15 4.02 rt 0.1 5 3.44 C 0.53 2.98 C0.11 1.78 0.03 * vc/w Vt/W 0.58 * C 0.02 0.35 k0.13 _f 0.42 k 0.09 0.41 f0.02 0.53 C 0.05 0.51 2 0.08 0.34 *** * 0.01 1.88 C0.19 1.21 * 0.03 1.61 * C 0.16 1.56 t 0.1 6 1.47 & 0.13 0.92 ** k 0.06 0.40 *** * * k 0.02 0.32 0.20 k 0.01 C 0.01 natal lung growth were made by Short ('50, '52). Mainly interested in the fate of alveolar epithelium during growth he concluded that additional alveolar surface area was formed postnatally by a subdivision process of alveoli which he interpreted to be due to a purely physical factor: the distension of the lung (Short, ' 5 0 ) , a theory, which he confirmed for other species including m a n i n his subsequent paper (Short, '52). The same parameters were estimated by identical methods in kittens and cats by Dingler ('58) who described the formation of new alveoli after birth. Of particular interest for our study is the work of Neuhauser and Dingler ('62) and Neuhauser ('62) who investigated the development of the inner surface of the lungs of suckling rats. They found a n irregular growth rate with a slightly steeper increase of alveolar surface area during the second week. For this period they also described the formation of new alveoli. Similar findings have been reported by Weibel ('67a). Using more elaborate morphometric techniques his study revealed that the important structural alterations of the rat lung took place within the first ten days of life. It was hence this short period on which we focussed our interest. *** Sa/W sc/w 0.531 C 0.029 0.371 C 0.040 0.471 +- 0.018 0.523 k0.018 0.499 C 0.041 0.559 C 0.061 0.277 +- 0.009 0.172 C 0.006 0.592 f0.050 0.326 C 0.093 0.432 2 0.058 0.484 f0.020 0.519 f0.034 0.550 C 0.068 0.291 & 0.008 0.179 & 0.005 * * *** *** * *** *** B. Discussion of some methodological aspects Because of the amount of work involved our investigation had to be based on the analysis of three animals per age group. Although morphometric techniques are rather accurate, it appears on closer inspection of our results, that for some groups and some parameters the standard errors are high. As a consequence some relatively large changes do not become statistically significant. This applies in particular to the capillary parameters on day 4 and 7, despite the fact, that for these ages six blocks per animal instead of three were evaluated. The most probable reason for this inhomogeneity is biological variation. Since this period is marked by rapid transformations, differences in the capillary development between the animals of one group could easily occur. We can, however, safely assume that the trend observed is not due to chance: for all species and all ages so far investigated with our techniques i t is known, that the capillary surface area (fig. 2) follows alveolar surface area rather closely (Burri and Weibel, '71a; Geelhaar and Weibel, '71; Siegwart et al., '71 ; Weibel, '72). Furthermore Weibe1 ('67a) also observed a fall of S, on day 5. 724 P. H. BURRI, J. DBALY AND E. R. WEIBEL E E 2 0- E E \ 0” -E > 1008- 06- I- i? 0 4 Q a Q 0 z G cn . 02- ’.. ./’ D, = 0.0073 . W 3 LL & .lo0 .08$ 06Q 6 / 04- // f 3 a 0 .02- 2 4 6 810 40 60 80 100 BODY WEIGHT g 600 200 20 w Fig. 6 Morphometrically determined pulmonary diffusing capacity (DL) for oxygen plotted double logarithmically against weight ( W ) . Two growth phases with distinct d o metric relationships c a n be m a d e out (solid line). In a first phase (days 1 to 21) DL increases in direct proportion with body weight; in t h e second phase it increases a t a significantly less steep rate. The broken line represents the overall allometric function DL = 0.0123 . W0.*j. Dots are animal values; open circle represents an aberrant value of one animal of day 4, which has not been considered for the regression calculations. The question may be raised, why did we not count alveoli? In fact we did; we counted alveoli or what appeared to be “alveoli” in one animal per group.3 The absolute number and the numerical density of “alveoli” and the mean “alveolar” diameter are presented in table 8 and in figure 7 respectively. The two curves (fig. 7) indicate that between days 4 and 13 a rapid increase in density of “alveoli” occurred combined with a rapid fall in “alveolar” size. The interpretation is that the original air spaces had become sub- 3 At a magnification of 400 times on the Wild M501 automatic sampling stage microscope we estimated on ten fields per animal the volume density of all “alveolar like” air spaces VV.* (which is not identical with the previously used Vva) and the mean number of “alveolar” profiles NA=*’per field. From these parameters the number of “alveoli” per unit volume of lung parenchyma (numerical density of “alveoli” Nva*) could be derived usine the formula: 1 NA.* 312 Nva* = - B . ~ V”.” 112 The value of 1.38 for the shape coefficient 4 was used (Weibel and Gomez, 6 2 ) , assuming alveoli to be roughly spherical; NA=*was derived from NA.*’and represents the number of alveoli” per unit area of section. From the absolute number of “alveoli” and the absolute “alveolar” volume we calculated the sp$;;;:u” diameter D.* by assuming “alveoli” npg 725 MORPHOMETRY OF POSTNATAL LUNG GROWTH TABLE 8 Absolute number of “alveoli” Na* and derived mean “alveolar” diameter Da+, assuming “alveoli” to be of constant spherical shape (see t e x t ) . Measurements were made o n one randomly chosen animal per group Age in days 1 4 7 10 13 21 Da* Na* 106 P 2.59 4.73 10.46 18.38 25.76 33.82 54.3 51.5 39.3 34.7 33.3 37.7 divided. This is in good agreement with our other results; the absolute values given here however are certainly incorrect, since a n “alveolus” at birth is obviously different from a n alveolus of the adult lung. Thus, we would like to emphasize that it would be erroneous to deduce from this curve, that true alveoli are present at birth (Burri, in prep. ). N ~ , * NUMERICAL DENSITY OF Even if one is aware of these transformations alveolar counts are inaccurate in the growing lung: ( a ) In some stages it is difficult or even impossible to identify safely the alveoli in statu nascendi. ( b ) During development the alveoli change their shape from very shallow to rather deep excavations. The problem of correction for alveolar shape variations has not been properly solved so far. As shape enters the formula for calculation of alveolar number either in form of a shape coefficient (Weibel, ’63, see footnote) or in form of the mean tangent diameter (De Hoff and Rhines, ’61), any calculation not properly considering the shape changes will be misleading. C . Discussion of the results 1. Lung volume and light microscopic morphometry The pattern of lung volume growth with “ALVEOLI ” “ALVEOLAR” 01A M ETE R Oa‘. Y 58 18, 54 16. 14 50 12. 46 10 42 8 6 38 4 .34 2, 1 1 I 1 4 7 10 AGE I 13 21 IN DAYS Fig. 7 Numerical density (Nv,+) and mean diameter ( D a b ) of “alveoli.” Fall in D,* and rise in Nv.* reflect the rapid partitioning of primitive air spaces occurring after birth; the values are however not representative (see text). Each data point was obtained from one animal only. 726 P. H. BURRI, J. DBALY AND E. R. WEIBEL respect to body weight, especially the fall in the ratio VJW after day 10 is in agreement with the findings of others (Neuhauser and Dingler, '62; Neuhauser, '62; Weibel, '67a). Our data indicate that after day 10 lung volume increases as the function W".'O. From the light microscopic morphometric analysis i t appears that the relative amount of parenchyma does not change during growth. Since the value of Vv, for day 4 is the same as that for days 1 and 7, we can refute the possible objection that the lung could merely have been overinflated during instillation. The airways do not grow in proportion to the lung volume : their absolute volume increases 2.8 times, lung volume four times from day 1 to 21; from day 21 to 131 the airways enlarge 3.3 times and lung volume six times. 2. Electron microscopic morphometry Based on the quantitative analysis three growth phases can be postulated in postnatal lung growth: ( a ) a short phase of lung expansion, ( b ) a phase of intense tissue proliferation , ( c ) a long period of equilibrated growth. (a) Phase of lung expansion The results of all parameters investigated show that from day 1 to 4 the lung undergoes a n expansion of its air spaces. Almost the entire change in lung volume is due to the 87% increase of air volume, which causes the volume density of tissue and the surface densities of alveoli and capillaries to fall on day 4 (fig. 2 ) . Analysis of the data of Dingler in cats ('58) and Neuhiiuser and Dingler in rats ('62) reveals a n analogous trend of Sva.I n rabbits, Short ('50) found the percentage of air spaces lumen in the lung to increase rapidly from 68.6% on day 1 to 80% on day 5. All these findings indicate that initially after birth, lung growth follows chest growth by lung expansion. The fact that tissue mass drags behind, favors the assumption that this could be mainly a passive process. ( b ) Phase of tissue proliferation This phase is marked by high rates of increase i n tissue mass (Vt, Vs) and surface areas (Sa, Sc) and by a reversed shift in lung compartments compared to those of phase ( a ) . The surface densities (S,,, Sv,) show a period of increase, a fact, which has also been described for Sv. by Short ( ' 5 0 ) , Dingler ( ' 5 8 ) , Neuhauser and Dingler ('62), and Neuhauser ('62). The morphologic findings presented in a companion paper (Burri, in prep.) will show that this period is marked by a n extensive and rapid proliferation of secondary septa1 crests, leading to a subdivision of the primary air spaces. It is rather difficult to determine the duration of this proliferative phase, since it depends on the parameter taken as reference. Based on some preliminary findings of this study and considering especially the explosive tissue proliferation we have postulated that this phase was very short and lasted until day 7 (Burri, Dbaly and Weibel, '72). It is however apparent that certain parameters, capillary volume (Vc) and surface area (Sc) continue to increase markedly between days 10 and 13; the capillary loading (Vc/Sa) expresses this clearly. Thus evidently capillary development postdates tissue development. The alveolar and capillary surface areas increase steadily from day 4 u p to day 21, when plotted double-logarithmically against lung volume (figs. 8,9). The curve is flatter before day 4 and after day 21. Weibel ('67a) has described a similar steep increase of alveolar surface area between days 5 and 10 with a markedly lower increase rate before and after this five day period. Thus, the present study confirms his data with the exception that the high rate of augmentation lasts up to day 21. It is obvious that the growth of alveolar surface area which equals the 1.6 power of lung volume during this second phase implies some structural transformations, as is also apparent from the curve of the alveolar surface to volume ratio (fig. 3 ) . I n a lung growing by mere enlargement of its alveoli the surface area would increase in the proportion to the two-thirds power of lung volume. These structural alterations comprise not only formation of new septa but also 727 MORPHOMETRY OF POSTNATAL LUNG GROWTH I 02 I 04 06 10 40 60 LUNG VOLUME IN 0 3 20 I 10 20 y Fig. 8 Bilogarithmic plot of alveolar surface area ( S a ) and lung volume (VL) for 1 to 131 day old rats. Between days 4 and 21 the slope of the curve is very steep. The curve is flatter before and later on. The two solid lines represent the calculated regressions for days 1 to 21 and 21 to 131 respectively. their lengthening. Both of these processes will be responsible for higher rates of increase than those provided by a proportional expansion. The lengthening seems to be achieved partially by stretching and by a redistribution of the tissue mass which does not increase between days 13 and 21. This stretching is also indicated by the marked decrease of Vvt after day 13, the fall of the arithmetic and harmonic mean barrier thicknesses, and by the reduction of septal thickness, I n this context the absolute mass changes in the tissue components are interesting. Apart from epithelium the rapid mass increase stops between days 13 and 21, which is in accordance with autoradiographic h d i n g s (Kauffman et al., in prep.). The significant loss of interstitial tissue mass recorded in the third week is surprising. A possible explanation for this finding is that the freshly formed intercellular fiber mass and the ground substance could be condensed. Since i t appears from our results that the initial part (days 7-13) of the steep rise in Sa and Sc is due to tissue proliferation, whereas the second part of this steep increase (days 13 to 21) is probably due to a rearrangement of tissue mass, we conclude that the phase of intense tissue proliferation lasts until the end of the second week. ( c ) Phase of equilibrated growth From our data we can deduce that a period of more equilibrated growth starts somewhere between days 13 and 21. During this phase, the newly formed respira- 728 P. H. BURRI, J. DBALY AND E. R. WEIBEL SC 8000 6000 LOO0 N r 5 z = 095 2000 Q w (r a 1000LQL n 3 600 cn > (L Q, 2 400 a Q 0 200 0 02 04 06 10 2 LUNG VOLUME 4 IN 6 I 10 20 VL cm3 Fig. 9 Bilogarithmic plot of capillary surface area (Sc) and lung volume (V,) for rats aged 1 to 131 days. The very steep slope between days 4 and 21 is preceded and followed by flatter segments. The two solid lines correspond to the calculated regressions for days 4 to 21 and 21 to 131 respectively. The open circle represents an aberrant value of one 4 day old animal; it was excluded from the calculations. tory units are progressively enlarged, the quantitative relationship of their components however can still undergo some shifts After day 21 tissue mass is further increased but to a lesser degree than air space volume (tables 3, 5a). On the other hand, the volume of the blood compartment (capillaries and larger blood vessels) showed the greatest increase. Alveolar surface area was enlarged 3.6 times between days 21 to 131, the alveolar air compartment 6.2 times. If one assumed alveolar surface area to increase by simple enlargement, i.e., by proportional growth of the existing air spaces, the surface should increase by the twothirds power of air volume change, or by a factor of 3.4, a figure very close to that which we measured. The exponent of the allometric relationship between Sa and VL is 0.71, i.e., slightly higher, but not significantly different from 0.67, the 95% confidence interval extending from 0.57 to 0.85. Correspondingly, the alveolar surface density decreased after day 21 but reached the adult value already on day 44 (750 cm2/cm3; Burri and Weibel, '71a). Moreover, the alveolar surface to volume ratio (S/V). remained also unchanged between days 44 and 131; these findings thus would favour the assumption that new gas exchange surface could be formed as long as the lung continues to grow. In rats, lung growth seems to be possible up to the adult age, since in this species the thoracic cage does not ossify until late adulthood (Dawson, '27). Weibel ('67a) and Bartlett ('71) MORPHOMETRY OF POSTNATAL LUNG GROWTH postulated that new alveoli continued to be formed in adult rats. These authors found by light microscopic morphometry Sa to increase a s a function of VLo.'* and VLo.8' respectively. We must consider however that a higher rate of increase of Sa (as compared to the function VL0.67) could also be achieved by expansion combined with a n overproportional lenghthening of the interalveolar septa. Based on the available data this problem cannot be settled and certainly needs further clarification. 3 . Morphometric estimation of pulmonary diffusingcapacity The pulmonary diffusing capacity in growing rats had been found to increase proportionally to body weight ( Weibel, '67a). This is partially confirmed by our data since we found this to be true for the first three weeks of postnatal lung growth. One can, however, notice that between days 4 and 10 six out of eight points lie below the curve (fig. 6). Analysis of the calculations for DL shows that this is due mainly to small capillary volumes. This fact could perhaps point to the critical aspect of postnatal pulmonary growth: the capillarisation and the full deployment of capillaries i n the newly formed septa is the latest process involved in the morphological transformations (Burri, in prep.). The finding that after day 21 DLincreased as a function of W0.73reflected the smaller rate of increase i n lung volume which occurred with body growth during this phase. Siegwart et al. ('71) in a study on adult dogs of different strains (and hence body weights) described DL to grow as a function of W'.". I n interspecies comparisons Stahl ('67) found physiological DL to increase with W'.'8 whereas Weibel ('72) found morphometric DL to increase with W0.96. These findings are contrasting with ours. We must however emphasize that our data were obtained on growing animals of the same species. Therefore they probably cannot be directly compared to those obtained by the authors cited above. However that may be, it is important to note that in the rat, lung function, a s defined by the morphometric pulmonary diffusing capacity, is not seriously affected in the immediate postnatal period by the massive transformations of lung structure. 729 It should be emphasized however, that the transformations here described pertain for the rat lung; further work will have to show to what degree differences in maturity at birth may shift the period of transformation with respect to birth. ACKNOWLEDGMENTS The authors wish to thank Mrs. Marlise Witschi, Mrs. Silvia Stoller and Mr. Karl Babl for their skillful technical assistance and Mrs. Esther Hassler for the typing of the manuscript. We are indebted to Dr. J. B. Forrest for his reading of the manuscript and his suggestions. LITERATURE CITED Barnard, W. G., and T. D. Day 1937 The development of the terminal air passages of the h u m a n lung. J. path. Bact., 45: 67-73. Bartlett, D., Jr. 1971 Postnatal growth of the mammalian lung: influence of excess growth hormone. Respir. Physiol., 12: 297-304. Boyden, E. A,, and T. H. Tompsett 1961 The postnatal growth of the lung in the dog. Acta Anat., 47: 185-215. 1965 The changing patterns in the developing lungs of infants. Acta Anat., 61: 164192. Bremer, J. L. 1935 Postnatal development of alveoli in the mammalian lung i n relation to the problem of the alveolar phagocyte. Contrib. Embryol. Carneg. Inst., 25: 83-118. 1936/1937 The fate of the remaining lung tissue after lobectomy or pneumonectomy. J. Thoracic Surg., 6: 336-343. Broman, I. 1923 Zur Kenntnis der Lungenentwicklung. Anat. Anz., Erg. Heft, 57: 83-96. Burri, P. H. The postnatal growth of the rat lung. Part 111. Morphology. In preparation. Burri, P. H., J. Dbaly and E. R. Weibel 1972 Postnatal growth of the rat lung. Experientia, 28: 727. Burri, P. H., a n d E. R. Weibel 1971a Morphometric estimation of pulmonary diffusion capacity. 11. Effect of Poz on the growing lung. Adaption of the growing rat lung to hypoxia and hyperoxia. Respir. Physiol., 11: 247-264. 1971b Morphometric evaluation of changes in lung structure due to high altitude. In: Ciba Foundation Symposium on High Altitude Physiology. R. Porter and J . Knight, eds. Churchill Livingstone, Edinburgh and London, pp. 15-30, Chalkley. H. W. 1943 Methods for the quantitative morphologic analysis of tissues. J . Nat. Cancer Inst., 4: 47-53. Clemens, H. J. 1955 Untersuchungen uber das maximale Lungenluft-Volumen. Ein Beitrag zum Problem des postnatalen Lungenwachstums. Morph. Jb., 95: 447-517. Cohn, R. 1940 The postnatal growth of the lung. J . Thoracic Surg., 9: 274-277. 730 P. H. BURRI, J. DBALY AND E. R. WEIBEL Davies, G . , and L. Reid 1970 Growth of the alveoli and pulmonary arteries in childhood. Thorax, 25: 6: 669-681. Dawson, A. B. 1927 Further studies on the eoiphyses of the albino rat skeleton, with special reference to the vertebral column, ribs, sternum and girdles. Anat. Rec., 34: 351-363. DcHoff, R. T., and F. N. Rhines 1961 Determination of the number of particles per unit volume from measurements made on random plane sections: The general cylinder and the ellipsoid. Trans. AIME, 221: 975-982. Dingler, E. L. 1958 Wachstum der Lunge nach der Geburt. Acta Anat. Suppl. 30, 32: 1-86. Dunnill, M. S. 1962 Postnatal growth of the lung. Thorax, 17: 329-333. Emery, J . L., and A. Mithal 1960 The number of alveoli in the terminal respiratory unit of man during late intrauterine life and childhood. Arch. Dis. Child., 35: 544-547. Emery, J. L., and P. F. Wilcock 1966 The postnatal development of the lung. Acta Anat., 65: 10-29. Engel, S. 1947 The Child’s Lung. Arnold, London. 1953 The structure of the respiratory tissue in the newly-born. Acta Anat., 19: 353365. Geelhaar, A,, and E. R. Weibel 1971 Morphometric estimation of pulmonary diffusion capacity. 111. The effect of increased oxygen consumption in Japanese Waltzing mice. Respir. Phvsiol., 1I : 354-366. Glagoleff, A. A. 1933 On the geometrical methods of quantitative mineralogic analysis of rocks. Trans. Inst. Econ. Mineral., Moscow, 59. Gnagi, H. R., P. H. Burri and E. R. Weibel 1970 A multipurpose computer program for automatic analysis of stereological data obtained on electron micrographs. Proc. 7th Int. Congr. Electron Microscopy, Vol. I, pp. 443-444. Grab, W., and W. Miinker 1964 Tierversuche. I. Zucht und Pflege von Versuchstieren. In: Biochem. Taschenbuch. 2. Auflage, 2. Teil. H. M. Rauen, ed. Springer Verlag, Berlin-attingen-Heidelberg, p. 255. Heiss, R. 1923 Bau und Entwicklung der Wirbeltierlunge. Ergebn. Anat. Entw. Gesch., 24: 244-277. Kauffman, S. L., P. H. Burri and E. R. Weibel. The postnatal growth of the rat lung. Part 11. Autoradiography. In preparation. Koelliker, A. 1879 Entwicklungsgeschichte des Menschen und der hoheren Tiere. 2. Auflage. Leipzig, p. 868. Luft, J. H . 1961 Improvements i n epoxy resin embedding methods. J. Biophys. Biochem. Cytol., 9: 409. Neuhauser, G. 1962 Beitrag zur Morphogenese der Lunge. Anat. Anzeiger, Suppl., 111, 277284. Neuhauser, G., and Ch. Dingler 1962 Lungenwachstum im Sauglingsalter. Zeitschr. Anat. Entw. Gesch., 123: 32-48. Norris, R. F., T. T. Kochenderfer and R. M. Tyson 1941 Development of the fetal lung. Amer. J. Dis. Child., 61: 933-950. Reid, L. 1967 The embryology of the lung. In: Ciba Foundation Symposium on Development of the Lung. A. V. S. de Reuck and R. Porter, eds. J. and A. Churchill Ltd., London, pp. log-124. Reynolds, E. S. 1963 The use o f lead citrate of high pH as an electronopaque stain in electron microscopy. J. Cell Biol., 17: 208-211. Scherle, W. 1970 A simple method for volumetry of organs in quantitative stereology. Mikroskopie, 26: 57-60. Short, R. H. D. 1950 Alveolar epithelium i n relation to growth of the lung. Phil. Trans. Roy. SOC.(B), London, 235: 35-86. 1952 Aspects of comparative lung growth. Proc. Royal SOC.( B ) , 140: 432-441. Siegwart, B . , P. Gehr, J. Gil and E. R. Weibel 1971 Morphometric estimation of pulmonary diffusion capacity. IV. The normal dog lung. Respir. Physiol., 13: 141-159. Stahl, W. R. 1967 Scaling of respiratory variables in mammals. J. Appl. Physiol., 22: 453460. Tomkeieff, S. I. 1945 Linear intercepts, areas and volumes. Nature, 155: 24. Weibel, E. R. 1963 Morphometry of the human lung. Springer-Verlag, Berlin-Gottingen-Heidelberg. 1967a Postnatal growth of the lung and pulmonary gas-exchange capacity. In: Ciba Foundation Symposium on Development of the Lung, A. V. S. de Reuck and R. Porter, eds. J. and A.-Churchill Ltd., London, pp. 131-148. 1967b Automatic sampling stage microscope and data print-out-unit. In: Stereology. H. Elias, ed. Springer Verlag, New York, pp. 330-331. 1969 Stereological principles for morphometry in electron microscopic cytology: h t . Rev. Cytol., 26: 235-302. 1970 An automatic sampling stage microscope for stereology. J. Microscopy (London), 91: 1-18. 1970/1971 Morphometric estimation of pulmonary diffusion capacity. I. Model and method. Respir. Physiol., 11: 54-75. 1972 Morphometric estimation of pulmonary diffusion capacity. V. Comparative morphometry of alveolar lungs. Respir. Physiol., 14: 26-43. Weibel, E. R., and D. M. Gomez 1962 A principle of counting tissue structures on random sections. J. Appl. Physiol., 17: 343-348. Weibel, E. R., and B. W. Knight 1964 A morphometric study on the thickness of the pulmonary air-blood barrier. J. Cell Biol., 21: 367384. Wicksell, S. D. 1925 The corpuscle problem. Biometrika, 17: 84-99 ( p . 8 9 ) . 1926 The corpuscle problem; second memoir. Case of ellipsoidal corpuscles. Biometrika, 18: 151-172 (p. 166). Willson, H. G . 1922 The terminals of the human bronchiole. Am. J. Anat., 30: 267-295. 1928 Postnatal development of the lung. Am. J. Anat., 41: 97-122.