ABSTRACT
Larvae of Triturus cristatus survive unilateral pronephrectomy at the early caudal bud stage.
The remaining pronephros compensates and in the main the results confirm those of Howland (1921) on Amblystoma punctatum.
A comparison between control and operated specimens by an analysis of covariance showed that the pronephros of the pronephrectomized group had significantly higher values for all the measurements made, within the range studied (5·5–7·5 mm. nose to cloaca; 10–17 days after operation). The measurements include the means of: the nuclear population; total cell volume; total lumina volume; overall volume of pronephros; internal surface area of tubules; antero-posterior length of pronephros; volume of individual cell of pronephros. The slopes of the regression lines in the two groups demonstrate that in all measurements except that of individual cell volume, the pronephroi of the operated group grow faster than those of the controls. During development the individual cell volume significantly falls in the control pronephroi, in contrast to the operated group where the mean volume does not significantly change.
The duct of the compensating pronephros is expanded in cross-section along its entire length. The duct of the operated side atrophies from the anterior region towards the posterior, becoming a degenerate strand of cells, subsequently losing its lumen throughout its length.
INTRODUCTION
As Howland (1916,1921) originally showed and many other workers have confirmed, an amphibian larva can survive satisfactorily with one pronephros, but it cannot survive bilateral pronephrectomy (Swingle, 1919; Shimasaki, 1930a; and Cambar, 1944 a, b, 1948) unless, as shown by Cambar, the mesonephros has become well established and functional.
The remaining pronephros, when one has been removed, becomes noticeably larger in Rana sylvatica (Swingle, 1919), R. nigromaculata and Bufo vulgaris japonicus (Miura, 1930 a, b), Rana dalmatina and other Anura (Cambar, 1948), Triton alpestris (Machemer, 1929), and Amblystoma punctatum (Howland, 1916,1921; Detwiler, 1918, p. 520). Hiller (1931) has also studied the problem in the latter species from the functional aspect in parabiotic twinning. However, except for the quantitative papers of Howland and Hiller, both of which appear to suffer from insufficiency of data, there is no information of quantitative nature.
The following account is a quantitative histological analysis of pronephric compensation. The results support Howland’s conclusions that a remaining pronephros adjusts itself to the artificially created situation by compensating in a variety of ways.
MATERIAL AND METHODS
The larvae of Triturus cristatus carnifex (Spurway, 1953; Fox, 1955a) used in 1952 and 1953 were from the same breeding pair, ♀ 46 and ♂39, which had been presented to Dr. H. Spurway by Professor G. Montalenti of Naples in 1950. Only those larvae laid in the spring of 1953 were quantitatively analysed, but others of the 1952 crop were examined for general morphological features. Operations were performed at the early caudal bud stage (equivalent to Harrison stage 22–24 for Amblystoma punctatum) after decapsulation in full strength Holtfreter solution. The pronephros in urodeles arises from nephrotome material beneath somites 3 and 4 (Field, 1891; Howland, 1921; Fales, 1935), and the duct from material beneath somites 5, 6, and possibly 7 (O’Connor, 1938; Holtfreter, 1944). Just posterior to the presumptive branchial region, the pronephric rudiment can be seen as a small protuberance bulging slightly from beneath the epidermis. It is closely associated with the presumptive fore-limb region (Harrison, 1915; Detwiler, 1918). Operations were made with sterilized glass needles (Hamburger, 1950). A small incision was made over the pronephric blastema, the epidermis was folded back, and the pronephric rudiment was visible as a small, whitish glistening mass. The latter was then removed, care being taken to damage the surrounding tissue as little as possible. Even so, in many cases, the operation damaged the fore-limb blastema, and some of the pronephrectomized specimens at fixation showed stunting or even absence of the fore-limb on the side of the operation. The embryos were allowed to remain in full-strength Holtfreter solution for 2 hours to facilitate healing, and then they were transferred to Holtfreter solution diluted ten times, containing 0·1 per cent, sodium sulphadiazine. The base of each crystallizing dish in which individual specimens were kept was covered with a thin layer of gelatin. Controls were treated in an identical manner except that instead of removing a pronephros on one side, a surface wound was made, but the pronephric blastema lying beneath the epidermis was undisturbed. In all fifteen controls (30 pronephroi) and ten pronephrectomized specimens (10 pronephroi) were statistically analysed. Members in each group were killed in Smith’s fixative at two periods, the controls at 10 and 17 days after operation, the pronephrectomized at 10 and during a short period around 15 days after operation. Specimens were embedded in paraffin wax, sectioned transversely at 10μ, stained with Ehrlich’s haematoxylin, and counterstained with aqueous eosin. Other specimens were stained with Mallory and Heidenhain’s azan stains. The following measurements of the pronephroi of the two groups were compared: (a) nuclear populations; (b) volume of the total mass of cells; (c) volume of the lumina of the tubules; (d) overall volume of the pronephros (volume of the mass of cells plus lumina volume); (e) internal surface area of the tubules; (f) anteroposterior length of the pronephros; (g) calculated volume of the individual pronephric cell.
The measurements of each individual specimen were related to its nose-to-cloacal length, the latter being used as an index of development. In the controls the two pronephroi were analysed and the mean of the particular measurement in question was taken. In order to obtain the nuclear population of the pronephros, the first 10 sections, then every other section, and, finally, the last 10 sections, were analysed. Every nucleus of the pronephric tubules seen in the section by focusing up and down was counted. Each section not counted was considered to be the average of the section in front and behind it. The method was the same as that used to investigate nuclear populations in the pronephros of Trituras cristatus karelinii (Fox, 1955b). Corrections to the crude counts were applied to allow for nuclear length at right angles to the section thickness (Abercrombie, 1946). This nuclear length does not change during the range of development studied in either control or operated specimens; in both groups it is approximately 12 μ. In order to measure the volumes of the tubule cells and lumina, sections were microprojected at 480 diameters, and the areas of these components were measured by a planimeter. The first seven sections, then every fifth, and, finally, the last seven sections, were measured (amounting to approximately 40 per cent, of the pronephros). The measurements of the remaining sections were interpolated in the series. Internal surface area of the tubules in each section was obtained by measuring the internal length of line of the tubule walls with a map measurer. A check of the method of sampling amongst the sections was made in one control pronephros by comparing results from the usual method, in which 19 out of 43 sections (44 per cent.) were measured, with those from measurements on 31 sections (72 per cent.). Differences for the volumes of lumina were 3· 8 per cent., volumes of tubule cells 3 ·2 per cent., and internal surface area 1·7 per cent. The sampling would thus seem reliable enough. The volume of the individual cell of the pronephros was found by dividing the total number of nuclei into the total volume of the tissue. This is justified, as binuclear cells are rare. Nose-to-cloacal length was obtained by summing the total number of transverse sections from the extreme anterior end to the region where the pronephric duct opens into the cloaca. The range of time studied was from 10 to 17 days after operation; total lengths of the larvae within this range were between 10 and 15 mm. approximately, and nose-to-cloacal length 5 ·5 to 7·5 mm. approximately.
RESULTS
(a) Quantitative changes
When the two control subgroups, killed at 10 and at 17 days after decapsulation and wounding, are compared with each other, the pronephroi show considerable differences (Table 1). The mean (of both the right and left pronephroi) of the nuclear population significantly increases by 23 per cent., volume of the lumina by 78 per cent., and total internal surface area of the tubules by 55 per cent. The mean total cell volume does not significantly change. As the total cell volume does not increase and the nuclear population does, the mean volume of the individual pronephric cell falls significantly, by about 14 per cent. As a result of the increased volume of the lumina and the constant total cell volume, there is a significant increase in the overall volume of the pronephros of 23 per cent. The mean antero-posterior length of the pronephros would appear to shorten slightly, but this change is not statistically significant. In terms of nose-to-cloacal length, these changes are shown by the regression lines in Text-figs. 1–7.
Means with standard errors of various measurements on the four groups of control and operated animals

Regression lines of nuclear population of the pronephros, related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and the upper line represent the compensating specimens.
Regression lines of total volume of pronephric cells, mm.3, related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
When the changes with time of the pronephrectomized specimens are analysed (Table 1) all the changes which take place in the controls between 10 and 17 days are seen. In addition, there are increases in the means of the total volume of cells and the antero-posterior length of the pronephros. Again, the greatest increases are those of lumina volume (149 per cent.) and internal surface area (86 per cent.). Means of the nuclear population show an increase of about 38 per cent., overall volume of the pronephros of about 61 per cent., total volume of tissue cells of 27 per cent., and the antero-posterior length of 12 per cent. The mean volume of the individual cell of a compensating pronephros shows a fall of 7 ·5 per cent, during the range of development studied. This, in contrast with the control groups, is not significant. In terms of nose-to-cloacal length, these changes are shown by the regression lines in Text-figs. 1–7.
Comparison between the control and unilaterally pronephrectomized groups of animals was made by an analysis of covariance to allow for the variation in tadpole length. The analysis showed that the pronephrectomized group had significantly higher values for all the measurements made (Table 2). Table 3 shows the absolute size of the means of these measurements in the control and in the pronephrectomized groups when adjusted for the size of the animal. Means are related to 6·33 mm. nose-to-cloacal length (overall mean of the two groups combined). Percentage superiority of pronephrectomized over the control larvae was as follows: means of–the nuclear population 17 per cent., total cell volume 34 per cent.; total lumen volume 100 per cent.; overall volume of pronephros 52 per cent.; internal surface area 49 per cent.; antero-posterior length 14 per cent.; and mean volume of the individual cell 15 per cent.
Analysis of significance, regression, and correlation in the various measurements between control (contr.) and unilaterally pronephrectomized (exp.) specimens. Thirty control and 10 experimental pronephroi

Absolute size of group means when adjusted for size of animal (nose to cloaca). Overall mean nose-to-cloacal length for both groups combined, 6 ·33 mm.

Further analysis demonstrated that there is a significant difference in the slopes of the regression lines of the two groups in all measurements, except in those relating to individual cell volume. It is apparent that in all measurements except the latter, the pronephroi of the operated animals grow faster than those of the controls.
Regression lines of total volume of the tubule lumina, mm.3, related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
Regression lines of overall volume of the pronephros (volumes of cells and lumina together) related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
(b) The pattern of morphological change
The numerical analysis of changes which take place during control and compensating development is the most satisfactory way of demonstrating morpho-logical change not always obvious from casual microscopic examination. Some features are particularly noticeable, however. The tubules of the unilaterally pronephrectomized group are large and inflated (Plate, figs. B, D) though the amount of swelling is variable in different specimens. Tubule walls are usually thinner than those of the controls (Plate, fig. A), although it must be remembered that the tubules in the compensating specimens not only dilate excessively, but also increase their tissue substance, which would help to counteract thinning of the tubule wall. The duct of the remaining pronephros is in practically every case expanded in cross-section right back to the cloaca (Plate, figs. C, E, F). This is in agreement with the findings of Howland (1916, 1921), Miura (1930 a, b), Shimasaki (1930b), and Cambar (1947). Simultaneously there are characteristic changes in the duct on the operated side. Of 43 operated specimens examined, 31 showed a duct atrophied in the anterior region to a mere strand of cells. This continues posteriorly as a horizontally flattened duct lying between the somite and yolk mass, and opening into the cloaca (Plate, figs. C, E, F). Twenty-two of the 31 specimens were of nose-to-cloacal length of less than 6 ·5 mm., and 9 were greater. Three of the younger specimens showed a tiny lumen for a short distance at the extreme anterior end. Eleven of the 43 operated specimens showed a strand-like duct completely atrophied along its length, though delicate connexions containing lumina were present joining it to the presumptive mesonephros. No lumen was present in the region where the duct joins the cloaca. Four of these 11 specimens were less and 7 greater than 6· 5 mm. nose-to-cloacal length, so that they were somewhat older on the average than the previous groups. One further specimen (7·2 mm. nose-to-cloaca) had a well-developed duct with a well-formed lumen from the mesonephros backwards. This feature is probably a precocious formation of the mesonephric (Wolffian) duct developed from the pre-existing atrophied pronephric duct. We may conclude that the duct of the operated side atrophies after the extirpation of its pronephros, as found by Machemer (1929) and Van Deth (quoted by Woerdeman & Raven, 1946), and that this reduction proceeds from the anterior region posteriorly. This conclusion agrees with that of Miura (1930a), Maschkowzeff (1934), Shimasaki (1930a), and Cambar (1947, 1948), but does not confirm that of Howland (1921) who considered that reduction began posteriorly. Unilateral extirpation of the pronephros does not affect the glomus of that side according to Howland (1921), Shimasaki (1930b), Van Deth (see Woerdeman & Raven, 1946), and Cambar (1949). Machemer (1929), and Maschkowzeff (1934), in contrast, considered that the glomus was smaller on the operated side. In 43 pronephrectomized specimens of Triturus, 17 specimens appeared to show no difference in glomus size on the two sides (12 specimens were less and 5 greater than 6 ·5 mm. nose to cloaca); 7 specimens had glomi practically the same size or slightly, but probably not significantly smaller (3 specimens less and 4 greater than 6· 5 mm. nose to cloaca); 10 specimens showed the glomus on the operated side definitely shorter than that of the non-operated side (4 less and 6 greater than 6 ·5 mm. nose to cloaca); and 5 specimens showed the complete absence of a glomus on the operated side (3 less and 2 greater than 6· 5 mm. nose to cloaca). Four specimens were damaged sufficiently to render the examination inconclusive. Twenty-four specimens out of 39 would thus show no size differences between the two glomi. The results would perhaps suggest that shortening or absence of the glomus after extirpation of the ipsilateral pronephric blastema is due to damage or to the removal of the presumptive glomerular tissue. If all this presumptive tissue is removed at the caudal bud stage, there is no regeneration from the surrounding tissue. Another factor to be considered is that unilateral pronephrectomy if successful often leaves a large gap adjacent to the glomus. The gut and developing lungs frequently bulge into this space and may either damage the glomus or render interpretation difficult. This does not happen on the other side. The conclusions reached would appear to be in accord with those of Howland, Shimasaki, Van Deth, and Cambar. Howland claimed that anterior and posterior nephrostomial funnels are regenerated from the coelomic endothelium in many operated embryos, and Burns (1934) and Hiller (1931) reported blind nephrostomial funnels at the site of the pronephros after extirpation of the latter and its duct rudiment in Amblystoma. Regeneration is denied by Machemer (1929), Hiller (1931), Maschkowzeff (1934), and Van Deth (see Woerdeman & Raven, 1946). In many operated specimens of Triturus (not quantitatively analysed) similar vestigial nephrostomial buds were seen. This does not necessarily mean that regeneration has taken place. There is always the likelihood that there has been incomplete removal of all the presumptive pronephric tissue, a view originally suggested by Hiller. Nephrostomial funnel regeneration is thus not proved. Extra nephrostomes were never seen in a compensating pronephros. Two nephrostomes only, as in the normal condition, were always present.
Regression lines of internal surface area of the tubules, mm.2, related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
Regression lines of antero-posterior length of the pronephros, mm., related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
Regression lines of individual cell volume, μ3 related to nose-to-cloacal length in mm. The dots and lower line represent the control and the crosses and upper line represent the compensating specimens.
DISCUSSION
Howland (1921) measured a compensating pronephros fixed 9 days after unilateral pronephrectomy at H 30–32. One control specimen considered to be at the identical stage of development as the compensating specimen was used. She reported increases in the following measurements: cubic content of the mass of cells 63 per cent.; nuclear hyperplasia 16 per cent.; length of tubules 21 per cent.; and the internal surface area over 100 per cent. The actual measurements of the latter were 2· 037 sq. mm. (compensating pronephros) and 1·007 sq. mm. (control). Other measurements were comparative only. Hypertrophy of individual cells was also considered likely (p. 377,1921). Within the range examined, compensatory increases in Triturus would seem to agree with these results in Amblystoma. In both genera there are small increases in nuclear population and tubule length, and substantial increases in the total volume of tissue cells and internal surface area. Individual cell hypertrophy is confirmed. Howland, however, claimed that unilateral pronephrectomy retarded growth. This is not confirmed in Triturus; any differences in total length or nose-to-cloacal length between control and operated specimens were not significant within the range examined.
If it is assumed that the permeability of the skin to the influx of water does not change in early larval life, then a greater volume of water will enter a larva in a unit time as development proceeds. The surplus water has to be eliminated from the larva via the pronephros. Simultaneously with increase in larval size, physiological activities associated with the pronephric tubules (e.g. absorption and secretion of metabolic products across the internal surface membranes) will increase. Increase in pronephros size would thus seem to be related to the functional needs of the larva, and there is a high degree of correlation of tubule lumina volume and internal surface area with increase in larval length.
Although no actual measurements of body volume and surface area have been made of control and compensating forms, it is extremely likely that they do not differ significantly at similar stages of development. There is no reason to believe that the permeability of the skin to the influx of water differs in these groups, and, as the larvae were preserved during life in similar osmotic media, similar quantities of water will enter larvae of the same age. As similar quantities of fluid would have to be excreted to maintain similar overall larval volumes, then it can legitimately be assumed that twice as much fluid will pass through a compensating pronephros as through one of the pair of control pronephroi. In other words, the former will do twice the work of the latter. The characteristic changes which take place during compensation must of necessity be bound up with this increased functional activity.
ACKNOWLEDGEMENTS
It is a pleasure to thank Dr. H. Spurway for the supply of eggs of Triturus, Mr. M. Abercrombie for his critical interest and advice, Dr. K. A. Kermack and Professor P. B. Medawar, F.R.S., in whose Department the work was carried out. Photographs were prepared by Mr. W. Brackenbury, and Miss R. Birbeck assisted in preparing the manuscript. The investigation was generously supported by the Agricultural Research Council, England.