Journal of Genetics 48: 164-175 (1947)
THE PROBLEM OF NON-HEREDITARY ADAPTIVE MODIFICATIONS
(COINCIDENT OR ORGANIC SELECTION)
By V. S. KIRPICHNIKOV
Institute of Cytology, Histology and Embryology and Institute of Zoology,
Academy of Science, U.S.S.R., Moscow-Leningrad

Recent years witnessed the rise of renewed interest in the question of the interrelation of non-hereditary adaptive modifications and natural selection. Among the large number of papers published on this subject, those printed in Russian occupy a significant place. The problems discussed by the Russian scientists are of considerable importance in clarifying the theory of evolution. A review of these papers may be of general interest. In this connexion it may be in place to give a brief outline of the development of the problems concerned.

I. THE HYPOTHESIS OF COINCIDENT AND ORGANIC SELECTION

*Cit. after Baldwin (1902)
Towards the end of the last century Lloyd Morgan (1896), Osborn (1896, 1897). and Baldwin (1896,* 1902) simultaneously and independently of one another proposed a new hypothesis of the selection of coincident variations or, as Osborn and Baldwin termed it, of organic selection. This hypothesis was supported by Poulton, Wallace, Gulick, and can be summarized as follows:

Individual plasticity is the property of all organisms. According to Morgan and Baldwin plasticity is acquired in the struggle for life as the result of natural selection. According to Osborn it is an inherent quality of living matter, the property of protoplasm. Osborn found it impossible to explain in any other way the experimentally established ability for adaptation to conditions of life not only unusual for the existing forms, but for their ancestors as well.

The plasticity of organisms is expressed in the form of change in habits and of morphological, adaptive alterations. Habit alters the structure of the organism including such basic organs as the skeleton, the circulatory system, etc.

If individual adaptive modifications happen to be present, those among the numerous divers congenital variations will be favoured that are coincident with a given useful modification. Such variations will gradually accumulate. If the conditions that called forth the given plastic modification are continued for a considerable period, the accumulation of coincident variations will progress slowly, but inevitably.

According to Lloyd Morgan and Baldwin this accumulation is the result of the extinction of all variations going in other directions and the preservation of all coincident variations. Since they enhance non-hereditary change and supplement its shortcomings. From Osborn's point of view the accumulation of hereditary variations similar to modifications has nothing to do with selection, but takes place automatically since all other variations are eliminated.

The plastic (phenotypical) adaptations will be substituted by hereditary ones. Non-hereditary alterations of habit and structure acquire phylogenetic significance. As an illustration Lloyd Morgan refers to alterations that take place in bone structure. Certain individuals reveal an increased growth of bone structure as the result of increased function and of the influence of contiguous parts in an altered environment and with the acquisition of new habits. This may be qualified as a plastic modification. After such an alteration congenital variations leading also to an intensified bone development will no longer be eliminated. Moreover they will also be accumulated due to their usefulness since they enhance the plastic modification. The accumulating hereditary variations will finally effect a similar strengthening of the bone structure which had been hitherto the result of habitual use, that is, had been due to the plasticity of the organism.
CybeRose note: This is the pseudo-Lamarckian position. In fact, Lamarck wrote, "whatever the environment may do, it does not work any direct modification whatever in the shape and organisation of animals."
The fixation of plastic modifications does not imply the direct inheritance of acquired characters and the results of the exercise or desuetude of organs. It should be attributed to the automatic accumulation or selection of coincident variations, and it satisfactorily explains the parallelism between hereditary and non-hereditary variability which is often observed in the case of many animals and plants. Osborn and Lloyd Morgan emphatically stress the point that their hypothesis undermines the Lamarckian principle of evolution and destroys its basic arguments.

Baldwin points to the fact that organic selection not only stabilizes modifications but further expands plasticity in the same direction (see also Gulick, 1905). According to Baldwin organic selection attains first rate importance in evolution because of the existence of broad and complex accommodations of a general nature. Detto (1904) likewise refers to such modifications, indicating that selection creates comparatively non-specific reaction systems which account for the existence of a 'modification reserve' (see also Lukin, 1942). Thus Osborn's 'self-adaptation' in an unusual environment can be easily explained without resorting to the inherent plasticity of the protoplasm.

Gulick (1905) likewise supports the hypothesis of the selection of coincident variations and calls it 'coincident selection'. He particularly stresses the significance of the isolating influence of modifications which hamper free cross-breeding. Lately, this particular role of modifications has been discussed in works on ecological isolation among birds (Promptov, 1934a,b, 1936; Hogben, 1940), on biological races of insects (Thorpe, 1940), races of fish Kirpichnikov, 1933; L. S. Berg, 1934) and many others. Muller (1942) has analysed the question in detail and illustrated his arguments with examples of temperature modifications in Drosophila.

Huxley (1942) cites numerous cases in which there is a non-hereditary barrier for cross-breeding at the first stage of divergent evolution.

The hypothesis of coincident selection proposed by the early American authors was gradually forgotten in the years that followed. Baur (1922) and other well-known apologists talk of individual plasticity only as the result of natural selection. Turesson 1922) concludes that non-hereditary differences are substituted at the very earliest stages of the formation of new races and ecotypes by similar hereditary changes. Thellung (1930) refers to a corresponding replacement, and explains it by the fact that hereditary variations are more specialized and require less energy for the development of the character. All these remarks are of a casual nature and are not in any way connected with the views of Lloyd Morgan and his contemporaries. Only in 1935 did this problem once more become the object of discussion, particularly in Russian works. In his latest very interesting book Huxley (1942) lays much stress on the principle of Baldwin and Lloyd Morgan's organic selection.

II. 'INDIRECT SELECTION' AND THE FIXATION OF MODIFICATIONS
ACCORDING TO KIRPICHNIKOV

In a paper on the problem of race-formation in fishes (Kirpichnikov, 1933) and especially in succeeeding papers (1935, 1940) an attempt is made to give a general analysis of the origin and the significance of non-hereditary adaptive modifications in the formation of species and in evolution (the hypothesis of 'indirect selection').
*Further on in the text this capacity will be termed 'adaptability'.
The capacity to produce adaptive modifications* is the result of selection of viability in a fluctuating environment, in other words the selection of plasticity, which takes plate among all organisms. Thus, the ability of plants to alter the length of their roots, the number and size of stomata, and so forth, depending upon the moisture of soil and air, is historically the result of adaptation by means of selection to an ever-changing environment. These modifications spread throughout an entire population and may be repeated for a number of generations.

Not only are constant morphological changes often present and play an important role, being particularly essential to immobile plants (see also Schmalhausen, 1941), but small regulatory, functional and morphogenetic ones also. In higher forms of animals, habits (conditional reflexes) become particularly significant (see, for instance, Krushinsky, 1944. 1946).

'Growth' modifications in fishes serve as an excellent illustration of how completely one modification can effect an organism. Fluctuations in water temperature as well as in the quantity of food in the water lead to serious changes in the rate of growth and differentiation. This leads to changes in the size of the fish at maturity and in the final stages of growth. Changes in the interrelations between the rate of growth and differentiation also influence the process of maturation itself, the quantity of sexual products, the form and. proportions of the body. Finally, the number of 'quantitative' characters is changed (Hubbs, 1926. 1940; Kirpichnikov, 1933, 1940, 1943, 1945; Gabriel, 1944, etc.). Such an all-round reaction favours adaptivity to new conditions and opens the road for future adaptation to changing environment.

The existing parallelism of hereditary (systematic) and non-hereditary variability in nature points to the significance of the latter in speciation. There are many cases when closely related species and subspecies of plants and animals differ inherently from one another by characters that are easily altered by external conditions.

The stabilization of non-hereditary variations in speciation is the result of selection. and not due to the direct inheritance of acquired characters. The relation between the organism and its environment, which is unbalanced when the latter is altered, becomes readjusted through selection. Thus 'the curves of adaptivity' are moved in the direction of the changing environment. These curves express the alterations in viability of the individual observed when there are fluctuations in certain ecological factors in nature.

Selection in this case is most intensive when it concerns the less plastic but the most vital structures of the organism. Organs or characters that can be most readily altered are least subject to selection. Nevertheless the course of their development and their correlation with other characters changes more or less through selection. As a result their Capacity to adapt themselves to former conditions, which have by this time become rare, disappears altogether.

This gradual destruction of a part of the original adaptability signifies the fixation of non-hereditary differences. The cause of such fixation lies in the disintegration of reaction systems and correlations in ontogeny, through the selection of viability in an altered environment. This disintegration is not the result of an automatic accumulation of mutations that are now uncontrolled by selection as Schmalhausen (1938, etc.) supposes, but is due to the selection of favourable mutations which pleiotropically destroy unnecessary correlations.

Xerophilous plants may serve as an example. Let us assume that a change in the number and structure of the stomata corresponds to the new conditions of a lowered humidity, However, like all adaptive modifications, these alterations are not perfect. Selection continue to change the structure of the plant: its leaves and roots, the intensity of metabolism, the biochemical constants of its plasm, etc. The entire development of the plant will be altered, so that when the former conditions of humidity return, the restoration the original structure and number of stomata will be impossible. The longer the selection progresses in the new environment the more probable will be such partial fixation of adaptive modification.

According to Tachtadgjan (1941) reversions in plants never lead to the complete restoration of the original ancestral form. The reversion results in a structure similar those of homologous organs of existing plants. Thus, the carpels of several species when reversions take place do not resemble the leaves from which they originated, but they are like the leaves of the existing form which differ considerably from the original type. Each organ undergoes a hidden 'latent' evolution paralleling the easily noted phenotypical evolution. This 'latent' evolution is based upon correlations in the individual development of homologous organs. Tachtadgjan's conclusions are an illustration of the possible means of fixation of non-hereditary differences through selection when there is a correlated change of several organs. Latent evolution does not only alter the 'hereditary base' as Tachtadgjan postulates, but it also leads to changes in development, in morphogenetic conformities, and, therefore, to alterations in the adaptability as well; the result of this process will be the gradual fixation of the modification.

All this concerns only the adaptive differences between diverging groups and not the adaptive characters themselves. In a changing but not as yet stabilized environment the diversities may become hereditary, but the capacity for modification will not necessarily be narrowed, it may only be transformed. The dependence of development upon external conditions will not necessarily be lessened. In other words, the partial or complete fixation of the differences between populations is possible: (1) when the fluctuations in environment are increased; (2) when the same degree of fluctuations in the natural conditions is preserved even though the environment changes; and (3) when the fluctuations of the environment are decreased. The latter condition acts, especially when a species disintegrates into smaller and more or less isolated populations.

According to the hypothesis of 'indirect' selection the process of adaptation may often be divided into two stages. The first stage leads to the intensification of an adaptability through selection, and the second to the perfection and fixation of the non-hereditary adaptive differences as a result of further selection..

'Indirect' selection in the sense of Kirpichnikov is the selection of variations which are more or less similar phenotypically to adaptive modifications. The modified organ alters by selection very slowly. In other words we really deal in this case with coincident.selection. The author therefore is of the opinion that it would be appropriate to substitute the term 'indirect selection' by the older term 'coincident selection' the more so since Darwin (1859), Schmalhausen (1938) and many others use the word 'indirect selection' to signify correlated changes (by means of selection) of favourable and useless characters connected with each other in development.

III. THE REPLACEMENT OF NON-HEREDITARY CHANGES BY HEREDITARY ONES
IN EVOLUTION ACCORDING TO LUKIN, GAUSE AND WADDINGTON

Similarly to Turesson (1922) and particularly Thellung (1930) Lukin (1935, 1936) that adaptive modifications are rapidly substituted by phenotypically similar mutations. This substitution is due to the greater selective value of the latter. Lukin (1935, 1936 see also 1939, 1940, 1942) notes as such the following: (1) the readiness of an organ at the time of the beginning of its function and its complete formation at a given time, (2) the independence of development from accidental fluctuation in the environment, and (3) a greater specialization of hereditary adaptations. Lukin finds that this conception is supported by the early appearance of hereditary differences between populations in natural conditions (1939, 1940) and by the progressive automatization of individual development Lukin makes no difference between the two phenomena observed in nature, namely: (1) the fixation of diversities or the elimination of the possibility of returning to the original state without compulsory automatization of development (coincident selection), and (2) the fixation of the character, the diminishing of its dependence upon the environment, its stabilization, and the growth of the role played by the autoregulatory processes in development (stabilizing selection).

In his latest paper Lukin (1942) suggests a classification of various types of adaptive modifications. Of particular interest are those modifications that are not adapted to the conditions responsible for their appearance, but happen to be important links in development. Such are the alterations in the sexual glands of animals, the development of which depends upon light, temperature, the landscape, and the presence of individuals of the opposite sex (Mashkovzeff, 1940; Svetozarov & Straich, 1941). Such are the manifold changes in the stage development of plants, diapause phenomena and many other adaptations. As a good example Lukin points to the development of parasitic worms that have two hosts. The completion of their life-cycle depends upon the succession of sharply I contrasting conditions.

In the same paper the author discusses the possibility of the substitution of nonhereditary variations by hereditary ones. Modifications adapted to periodically fluctuating factors of the environment are seldom substituted by mutations. The preservation of such lability is also displayed by the already mentioned basic alterations in development which depend upon external influences. A twofold interaction with the environment, when the active and the selective factors do not coincide, is usually extremely valuable and represents one of the most perfect forms of adaptability.

In fact only those alterations are substituted which are the result of a constant or one sided change in the environment accompanied by the stabilization of the latter.

Gause (1940a, 1941a), in general agreement with Lukin, discusses the question of the replacement of adaptive modifications by mutations in the course of selection. He support his views by the results of a number of experiments on selection among Infusoria. He calls this selection substituting or stabilizing selection (Gause, 1941a; Gause, Smaragdova & Alpatov 1942). notes that in his and Smaradova's (1940) experiments there was not a complete phenotypical likeness between adaptive modifications and mutations, and that their similarity is only partial. Later Gause (1941b) discusses the problem of ecological adaptability and classifies these phenomena from the physiological point of view.

Discussing the same problems in regard to acquired and inherent calluses Wa addington 1942) introduced the new idea of the canalization of individual development. The ability to react develops under the controlling influence of natural selection. In later stages, as a result of further selection, this ability becomes canalized. This signifies that development may progress only in one or a few directions due to the regulation against accidental deviations and the perfection of threshold reactions. External stimuli act as 'arrows'; the differentiation of the intermediate type becomes impossible. The selective value of the complete independence of the end product from the environmental fluctuations during development of an organ finally leads to the substitution of such 'arrow' reagents by genetical ones when the latter begin to act earlier in the development.

IV. THE VIEWS OF SCHMALHAUSEN, MURETOV AND MASHKOVZEFF

Schmalhausen (1938, 1939, 1940, 1941) mentions three means of hereditary fixation of non-hereditary, diversities: (1) fixation in the process of 'indirect' or coincident se1ection Kirpichnikov, 1935) which proceeds very slowly since favourable mutations on which this selection is based are very rare, (2) fixation due to the uncontrolled accumulation of numerous genic variations, which destroys the ability of those adaptive reactions which are no longer supported by selection; such an automatic destruction of the unutilized ability progresses with greater rapidity since it is connected with the accumulation of numerous unfavourable mutations, and (3) fixation in the process of the all-significant 'stabilizing' selection. This is the most rapid and the most significant means of fixation in evolution.

The theory of stabilizing selection (see also Schmalhausen, 1946) is deduced from the assumption of a high rate of mutations in nature and of their unfavourable character. The preservation of the existing adaptive normal phenotype is possible only on condition that the mechanisms of heredity and development are continuously readjusted.

Selection tends to preserve the normal type despite a high mutability which denotes hat new genetical combinations are tested and that the most stable survive. Development thus changes and becomes less dependent upon external influences, although the phenotype of the mature form is not altered.

Such selection implies the general growth of the stability of development, the perfection of various defensive reactions with the growth of antoregulation. Regulatory processes become characteristic for development in general, particularly in higher animals. Reactions connected with the moulding of organs become 'threshold' ones (Goldschmidt, 1938). The distance between the extremes within which the given reaction becomes possible, is increased (Filatov, 1939). Thus development becomes better protected against all kinds of violations (Schmalhausen, 1938, etc.; Muretov, 1941; Kamshilov, 1941; Waddington, 1942). The direct mechanical influence of adjacent parts upon one another (Filatov, 1941) may acquire regulatory significance. The same may be said of processes connected with Child's physiological gradient (Kamshilov, 1941) and the various physiological systems which provide for the adjustment of functions (Kirpichnikov, 1940). Regulatory pheno-mena, particularly the intensification of sensitivity of organisms (Kalabuchov, 1940a,b) compensate their seeming emancipation from the environment.

Stabilizing selection leads to the decrease of variability due to the regulation of all kinds of minor changes and the growth of the influence of internal factors in development. The idea of correlated evolution of such principal properties as the stability of development, adaptability, mutability and dominance were concisely postulated by Berg (1944).

Stabilizing selection according to Schmalhausen (1938, etc.) is the principal integrating factor of evolution perfecting development and establishing morphogenetical correlations.

Stabilizing selection acts in all instances, notwithstanding the absence or presence of adaptive modifications. In case the last make themselves felt in the processes, stabilizing Selection gradually secures the structure of the organ which had been modified. The mechanism of fixation in this case is reduced to the accumulation of mutations and their combinations within the limits of the new adaptive phenotype. The genotype and the development of the organism are likewise reconstructed in correspondence to the new phenotype. In other words, external factors of development are gradually substituted by internal ones. As a result a new and better adapted phenotype is obtained with a maximally stable and independent development.

The substance of the theory of stabilizing natural selection in contrast to 'direct' natural selection is that the former takes place on the basis of selective advantage of the normal phenotype in comparison with the deviations (Schmalhausen, 1941, p. 315). According to Schmalhausen, the selection of favourable deviations from the normal type does not occur in this case. Since, however, Schmalhausen (1941) agrees that the development of the organism is perfected and the role of regulation in its development is increased as a result of which a better adapted phenotype is achieved in the above-mentioned case, the author of these lines does not see the necessity of drawing a strict line between the two conceptions (see Kirpichnikov, 1944b).

Experiments on Drosophila have shown that small harmful physiological mutations are extremely frequent (Dobzhansky & Queal, 1938; Muretov, 1939, 1941). Muretov speaks of the selection of viability which continually introduces manifold minor genic variations into the normal genotype without a noticeable change in the phenotype. Only the amplitude of reaction in the case of a number of characters is changed. If adaptive modifications happen to be present they become fixed. Thus Muretov's point of view is very similar to that of Schmalhausen's.

The principal evidences in favour of stabilizing selection are the following (Schmalhausen, 1941): (1) progressive development of regulatory mechanisms in evolution, (2) the parallelism of hereditary and non-hereditary variability, (3) incomplete 'reversion' of former characters when the original conditions return, that is partial inheritance of differences between the lower taxonomic categories, (4) the success in elimination of harmful effects of mutations in laboratory strains due to the selection of the most viable combinations of genes, (5) the constancy of the 'wild type' in natural populations in spite of a high mutability, (6) the dominance of the 'wild type', that is the regulation of heterozygous mutations as a partial result of the stabilization of development, and (7) direct experiments on selection carried out by Gause.

Both Schmalhausen (1941) and Lukin (1942) consider the perfection of regulatory mechanisms in the course of evolution due to which development becomes more autonomous and external factors become substituted by, internal ones, as an important proof of the correctness of their point of view. Mashkovzeff's (1936) experiments illustrated this substitution. He compared the ontogenesis and the role of function in the development of nephric ducts, gills and lungs in various representatives of vertebrates.

In agreement with Roux (1881) Mashkovzeff distinguished three stages in the evolution of ontogenesls. In the firsts stage the leading role in development should be attributed to the environment and to the function of organs. Secondly, the role of function is decreased, the influences of chemical factors and hormones in the embryo itself predominate; the earliest periods of development depend upon inner, hereditary stimulants (all vertebrates including mammals). The third stage is that of labile determination and self-differentiation characterized by complete autonomy of morphogenesis; the development of organs is achieved without the participation of function (partly birds, mostly insects).

The gradually decreasing morphogenic role of function in the course of evolution and, thus, the stabilization of development, as shown by Mashkovzeff'a experiments, cannot serve as evidence of the action of stabilizing selection in favour of the fixation of adaptive modifications. These facts as well as some of the points made by Schmalhausen only prove the existence of an important evolutionary tendency towards progressive stabilization of development. In order to verify these conceptions it is necessary to prove them by experiments in selection.

V. THE EXPERIMENTAL CONFIRMATION OF THE HYPOTHESES OF COINCIDENT
AND STABILIZING SELECTION

(1) It is a well established fact that the mutability of laboratory strains of Drosophila is decreased and variability is in general lowered, whereas the dominance of the normal phenotype is increased. These phenomena are probably the result of selection in conditions of great isolation (Berg, 1942a). To this the author would add the consideration that laboratory conditions are characterized by maximum stability, in which variability loses its adaptive significance; besides which it is decreased by the interference of man. In this case stabilizing selection can take place.

(2) The weakening and elimination of the harmful effects of mutations when they arise in laboratory conditions (Fisher, 1928; Schmalhausen, 1938 and others) can also be the result of stabilizing selection.

(3) The experiments of Kamshilov (1939) on the increase and decrease of the effectiveness of the mutation 'eyeless' in Drosophila melanogaster permit the conclusion that the selection of individuals with the most notable development of characters results at the same time in the latter becoming more independent of the environment. In other words, the degree of dependence of the development of an organ upon fluctuations of the elements of environment (in this case upon the temperature changes) may be easily modified by means of selection. At the same time these experiments indicate a correlation between the degree of development of characters and their stability.
     Indirect data of a similar kind may be found in the work of Berg (1942b, 1945) on the correlated differences in mutability, non-hereditary variability and dominance among related populations of Drosophila. All these properties determine the general stability of development, i.e. the extent of dependence upon the environment.

(4) Applying X-ray's to a large number of Drosophila, Naumenko (1941) obtained non-hereditary changes in the wings of all individuals. By selecting the individuals with the most clearly expressed modifications, strains with the same, but already hereditary changes were obtained. Such artificial stabilizing selection shows the possibility of a similar stabilization in nature.

(5) Smaragdova & Gause (1939) and Gause (1940b) successfully subjected the Infusoria Paramaecium caudatum and Euplotes vannus to a series of experiments with selection in conditions of different salinity. The original cultures adapted themselves relatively quickly to increasing salinity by means of individual adaptability.
     Unfortunately, there was no analysis of the adaptive characters themselves in these experiments. The later works of the same authors with Paramaecium bursaria are more illustrative. Gause (1941a) has shown the adaptive value of decrease of size in Infusoria in high temperature as well as the opposite process in case of low temperature. Smaragdova (1941) found a similar hereditary adaptive geographical variation.
     In experiments where selection was practised (Smaragdova, 1940; Gause, Smaragdova & Alpatov, 1942) a hereditary increase and decrease in size was obtained. This change at first appeared—true, in a lesser degree—as a non-hereditary modification. At the same time a new adaptive character—a narrow body—was likewise obtained. These are the first experiments that prove the existence of coincident selection. The authors for some reason call it stabilizing selection, although the adaptability in the strains obtained was not less than in the original ones.
     Such are the scant experimental materials. Meagreness of investigations is explained by the difficulty of carrying out exact experiments on selection. There is no doubt that further experiments will follow in the future.

VI. SUMMARY. THE SIGNIFICANCE OF ADAPTIVE MODIFICATIONS IN EVOLUTION

Let us attempt to outline briefly the mechanism of coincident selection without any further reference to the above-mentioned literary sources.

1. Natural selection is the factor responsible for the existence of the individual plasticity of organisms.

2. This plasticity is expressed in the first place by the development of non-reversible morphological adaptive modifications when the environment changes, and secondly by regulatory processes and complex systems of easily reversible physiological changes including reflexes. Due to the regulation of the various changes induced by the environment, development becomes less dependent upon the surrounding conditions; but the seeming 'emancipation of the organism from its environment' is compensated by more perfect and diverse regulatory and functional adaptability.

3. Selection creates and perfects reactions of a complex type that are common among many different species of organisms. This facilitates the rapid and coordinated adaptation of the organism to the new environment. It explains the existence of the so-called 'modification reserve' and often attributes a leading role in evolution to adaptive modifications.

4. Adaptive modifications engender the isolation of consanguineous varieties when Interspecific divergence occurs.

5. The selection of partly coincident variations leads to the fixation of non-hereditary differences. This fixation is achieved through the creation of new correlatory mechanisms in the course of selection as well as through the destruction of superfluous ones. Thus adaptability is changed by selection. This denotes the readjustment between the organism and the environment, the mutual connexion of which was violated when the latter was altered.

6. Coincident selection is the selection of small physiological mutations: they acquire the greatest value in this phenomenon. Such alteration of the basic physiological characters is the result of direct natural selection. Thus the appearance of stable hereditary, differences in physiological characters at the earliest stages of divergent evolution finds its explanation. The direct selection of viability through physiological characters in the presence of modified organs leads to the partial fixation of. the adaptively modified structures. Therefore such selection happens to be at the same time coincident selection.

7. Adaptive modifications indicate the possible channels of evolutionary transformation, they accelerate and facilitate the latter. Coincident selection perfects non-hereditary adaptations.

8. Selection that stabilizes development, making it less dependent upon the environment, may be considered as one of the forms of coincident selection if adaptive modifications are present. Stabilizing selection is often but not always connected with the increase of the constancy of ecological factors. It increases the significance of internal factors in development. Coincident selection leads to the fixation of diversity between populations whereby the adaptability is by no means in all cases decreased. Stabilizing selection on the contrary leads to the fixation of characters to a certain degree.

9. The advantages of the surviving forms and not the automatic exclusion of deviations from the normal type are of decisive significance in stabilizing selection.

10. The evolutionary process as a whole is characterized by progressive stabilization of development, particularly in the earlier embryonic stages. Nevertheless this general tendency should not be mistaken for stabilizing selection, the existence of which may be proved only by direct experiments.

11. Stabilizing selection lowers the variability and increases general regulatory ability of the organism including the regulation of mutations. Regulation of heterozygous mutations signifies an increased dominance of the 'wild type'. When the dependence of development upon external conditions is increased, the results are just opposite of what has been said above. In some cases of coincident selection variability may be altered but it does not necessarily decrease or increase.

12. Coincident (organic) selection explains the parallelisms between hereditary and nonhereditary variability without adhering to the Lamarckian principle of the adequate transmission of acquired characters or to any other orthogenetical conceptions. This fixation can be easily explained by the usual Darwinian factors of evolution, namely variability, heredity, the struggle for life, and natural selection.

The author wishes to acknowledge his gratitude to R. L. Berg for kind assistance in supplying him with the necessary literature, advising him and editing this paper. Without this friendly help, the paper could not have been finished for publication, since the author serves in the ranks of the Army.

REFERENCES