The Origin and Development of the Nervous System

lower range of concentrations or intensities, also experi- mentally determined for each species, the rate and degree of acclimation or acquirement of tolerance varies directly with, though not necessarily proportionally to, the general physiological activity of the protoplasm, and the rate and degree of recovery after temporary exposure to the action of the agent also varies in the same way. The metabolic reactions and particularly the oxidations are important factors in the general physiological activity of protoplasm, and many lines of evidence show that susceptibility may be used, within certain limits and with certain precautions, as a rough comparative measure or indicator of the rate of oxida- tion. This of course does not mean that all the agents used in determining susceptibility act directly on the oxidations. Undoubtedly different agents act in very different ways upon living protoplasm, but the general nonspecific character of the susceptibility gradients demonstrated with many different agents indicates the interrelation of different processes and conditions in the protoplasmic system. The difference between a region of high and one of low rate of oxidation unquestionably involves not simply the oxidative reactions but many other factors, e.g., colloid dispersion, permeability of limiting surfaces, active mass of enzymes, etc., and within physiological limits change in the physiological state involves changes in all these and doubtless other factors. Apparently the general relation between sus- ceptibility and physiological state is primarily an expression of the fact that the dynamic equilibrium of the more active state is, on the one hand, more readily upset by any extreme external action than that of the

less active state and, on the other, undergoes more rapid and more complete recovery from temporary alteration. In short, so far as susceptibility is non- specific and quantitative, it is apparently an indicator of the quantitative aspects of physiological state in protoplasm. Susceptibility to the higher, directly lethal concentrations and intensities can be determined by survival time, and in many cases loss of motor activity, swelling or shrinkage of cells, or other changes preceding death can be used as a check on survival time. To lower concentrations, which are not directly lethal, sus- ceptibility can be determined by the degree of inhibition, acclimation, or recovery in growth, development, motor activity, etc., in different body regions or individuals.

The data on susceptibility have demonstrated the existence in a large number of organisms of definite susceptibility gradients as a characteristic, and as the earliest distinguishable features of axiate pattern, not only for the organism as a whole but for various axiate organs and parts. In all cases the differences in sus- ceptibility correspond to differences in structure, rate of growth, development, and differentiation, and in cer- tain cases it has been possible to show that susceptibility gradients correspond to gradients in oxygen consump- tion and carbon dioxide production.

In various cases it has been possible to follow the apico-basal or anteroposterior gradient from the egg and to show the modifications which occur during the course of development and the appearance of new gradients in particular organs, or in connection with agamic reproductive processes. In the simpler organ- isms the susceptibility gradients are wholly or largely superficial. In the ciliate infusoria, for example, axiation and morphological differentiation are, at least to a very large extent, confined to the ectoplasm, and in the forms which have been examined the ectoplasm alone shows a gradient in susceptibility (Child, 19146). As regards at least many plant cells the same is also true. In many of the monosiphonous algae with elongated cells the apico-basal susceptibility gradient is very distinct within the limits of a single cell (Child, i<)i6c, e, 19170, b, 19200) as well as in the axis as a whole. Modifications with advancing age, cessation of growth, budding, etc., appear very clearly in the changes in susceptibility.

The symmetry gradients have been demonstrated in various animals as susceptibility gradients. In cer- tain siphonophores, for example, in which the necto- medusa is bilaterally symmetrical in form, the body ectoderm shows a marked bilateral susceptibility gradient In the developmental stages of the sea urchin (Child, 19160) the symmetry gradients have been observed. In certain polychaete annelid larvae the susceptibility to lethal concentrations has been found to decrease from the median ventral region laterally and dorsally, and in various fishes (Hyman) and Amphibia (Bellamy, 1919) it decreases laterally and ventrally from the median dorsal region.

In some animals the primary gradients persist throughout life, while in others the original gradients may completely disappear during development. In the development of the hydroid, for example, a very distinct gradient appears in the unfertilized egg and cleavage stages, and the planula shows a well-marked apico-basal gradient. But as the time of attachment of the planula approaches, this gradient gradually becomes less and less distinct, and the planula attaches itself by the end which was originally apical, a fact which has long been known. After attachment the original gradi- ent disappears completely, a new gradient in the oppo- site direction arises at the opposite end, and the first hydranth arises from the high end of this new gradient. In other words, the phenomenon which Loeb and others have called heteromorphosis is a normal feature of hydroid development. Again, in the polyclad Turbel- laria the early stages of development show a very distinct anteroposterior susceptibility gradient, the head arising from the most susceptible region, but in the later stages, so far as they have been examined, a reversal in the susceptibility relations occurs, at least in the superficial regions, and in the adult worms the ectoderm of the head is less susceptible than more posterior levels. These data on hydroids and polyclads have not yet been published in full. In the annelids a second gradient in the opposite direction from the primary antero- posterior gradient appears, at least superficially, in the posterior regions as the result of the development of the posterior growing region (Hyman, 1916; Child, 191 7 d). But, whatever the changes, they are definite and orderly and associated with the course of develop- ment in each species.

As development and differentiation progress the indications of more or less specific relations between particular regions or organs and particular agents become more frequent. In Planaria dorotocephala, for example, we find that the lateral margins of the body are more susceptible superficially than in the median regions to alkalies, while in neutral and acid agents there is little difference between median and lateral or the median ventral region is slightly more susceptible than dorsal and lateral. With methylene blue of cer- tain concentrations Mr. McArthur has found that susceptibility of the median ventral region is distinctly greater than that of dorsal and lateral regions. The early embryonic stages of Planaria have not been available for work on susceptibility, but the data on other flatworms and various other bilateral inverte- brates indicate that primarily the susceptibility of the median ventral region is higher than that of dorsal and lateral regions and the outgrowth of tissue from cut surfaces in pieces of adult Planaria suggests a greater parenchymal activity in the median ventral region. In the previously published data on the gradients in Planaria (Child, 1913^) it was noted that the relations as regards median, lateral, and dorsal were not entirely clear. As regards the alimentary tract, various facts indicate that in the adult the region of greatest activity and highest susceptibility is in the middle of the body about the base of the pharynx and that a gradient of decreasing activity extends in both directions from this region, but since the alimentary tract is an internal organ and is not readily separable from other parts it is difficult to obtain conclusive evidence on this point. The anteroposterior gradient of ectoderm and body wall persists throughout life, but with the appearance of new zooids at the posterior end new gradients arise in that region, or, more strictly speaking, the original gradi- ent undergoes modification. Apparently the primary

embryonic relations have undergone more or less alter- ation, even in Planaria. With the appearance of more or less specific relations between particular regions or organs and particular agents, the value of the susceptibility method as a means of distinguishing general quantitative differences and relations is of course greatly decreased. In the higher animals apparent specificity of relation between particular organs and particular agents is much more evident than in lower forms and becomes increasingly complex with the progress of differentiation, but even in these forms the general nonspecific susceptibility relations appear in the earlier developmental stages, at least in all forms examined.

While caution is always necessary in interpreting the data of susceptibility in nonspecific quantitative terms it has become more and more evident as the data have accumulated that general nonspecific susceptibility relations do exist, particularly in the simpler organisms and in the earlier stages of development, and that they are indications of fundamental physiological features of organismic pattern. It has also been possible to control and modify development in definite predictable ways through the differential susceptibility of different levels of the axial gradients.' Such modifications consist in differential inhibition, differential acceleration, differential acclima- tion and differential recovery, each representing a definite teratological type.

In cases of differential inhibition with respect to a gradient the degree of inhibition varies directly with the susceptibility and activity of different levels. The region of greatest activity is most inhibited and less active regions are less inhibited. In the apico-basal or anteroposterior axis, for example, the degree of inhibi- tion is greatest in the apical or head region and decreases basipetally or posteriorly, consequently the positions and proportions of parts are altered in a definite way, the apical or head region being relatively smaller and the basal or posterior regions relatively larger than in the normal form. Microcephaly, for example, is a characteristic result of differential inhibition along the polar axis. In differential acceleration the altera- tions of proportion are in the opposite direction, and megacephalic forms result. In differential acclimation and recovery growth or development is first inhibited to some extent, but the more active levels of a gradient acclimate or recover more readily and more completely than the less active, so that in these cases also growth or development is finally relatively more rapid or greater in amount apically or anteriorly than basally or posteriorly.

Similar modifications also appear with respect to the symmetry gradients. In differential inhibition in bilateral forms, for example, median regions are more inhibited, while in differential acceleration they are more accelerated than lateral, and in differential accli- mation and recovery median regions are finally less inhibited than lateral. These various modifications are, as regards their more general features, nonspecific as regards agents, and all except the differential accelerations which require the action of accelerating agents can be produced in some degree by a large number, probably by all agents which inhibit general protoplasmic activity and which

in lower concentrations or intensities permit at least some degree of acclimation or recovery. Moreover, modifications of the same general type with respect to a particular axial gradient can be produced in widely different organisms, e.g., flatworms, echinoderms, fishes, frogs. In many respects these definite developmental modifications constitute the strongest evidence for the relation between susceptibility and the rate of funda- mental activities of living protoplasm.1 And finally, differential susceptibility, as a relation dependent primarily upon quantitative rather than upon specific or qualitative differences in the physiological state or activity of protoplasm, provides a simple and adequate basis for the interpretation of much of the work on experimental teratogeny and many of the teratological forms occurring as " accidents" in nature. The cases of cyclopia and microcephaly in fishes, experimentally produced by Stockard (1907, 1909, 1910, 1911, etc.), are essentially similar to the inhibited types of head in Planaria (Child, 191 ia, 191 5^, pp. 105-17) and to the differential inhibitions in the sea urchin (Child, 1916 d) and in Amphibia (Bellamy, 1919, and further data not yet published) . All these cases involve a greater degree of inhibition of apical or anterior and median as com- pared with basal, posterior, and lateral regions and all are nonspecific in origin, i.e., can be produced by the action of various agents and conditions.

The susceptibility method makes no pretense of being an exact quantitative method of measuring 1 Data on the control and modification of development through differential susceptibility have appeared as follows: Child (19114, 1916^, 1917^); Bellamy (1919), and further data on Hydrozoa, echino- derms and Amphibia are still unpublished. metabolism or oxidation, nor is it to be regarded as taking the place of any other method of investigating physiological condition or rate of metabolism or oxida- tion. Its chief value is as a supplement to other methods. In the first place it enables us to demon- strate the existence of certain characteristic, non- specific regional differences in physiological condition in organisms, which because of their unicellular character, or their small size, or because of the complications intro- duced by separating different body regions, are not available material for other more direct and more exact methods. Even in these forms, however, the action of external chemical agents is in general from the surface inward, consequently the information given by the sus- ceptibility method concerns, first of all, the superficial regions of the cell or body, but it is possible in many cases to learn something concerning differential suscepti- bility of internal parts and organs.

Second, by the modification and control of develop- ment through differential susceptibility, the method enables us to show that the differences in condition indicated by differences in susceptibility are fundamental factors in organismic pattern. The conclusions con- cerning the relation between susceptibility and rate of metabolism or oxidation mean no more than that rate of metabolism or oxidation is a factor in the conditions which determine susceptibility, and it is not claimed that this conclusion is universally valid. Undoubtedly, in the more highly differentiated organisms and in more advanced stages of development the qualitative differ- ences in different organs may determine differences in susceptibility which are more or less specific as regards

both organ and agent, but in the simpler forms and the earlier stages the similarity of the susceptibility gradients in widely different organisms and with a great variety of agents renders their nonspecific character sufficiently clear, and many lines of evidence, both direct and indirect, indicate their relation to rate of metabolism or oxidation. The nature and degree of that relation in each particular case and for each particular agent remains of course to be determined by other methods of investigation. The susceptibility method has served to bring to light certain characteristic features of organismic pattern which have not previously been clearly recog- nized, viz., the gradients, but conclusions concerning the exact nature of these gradients are possible only on the basis of all the different lines of evidence obtainable, and at present of course cannot be final.

In many forms, both plants and animals, in which susceptibility gradients exist, corresponding gradients in the rate of penetration of certain substances, particu- larly the vital dyes neutral red and methylene blue, which have been most extensively used in these experi- ments, have also been demonstrated.1 The existence of these gradients in rate of penetration raises the question whether the susceptibility gradients are not primarily gradients in permeability of the protoplasmic surfaces to the agents used. While there is no doubt that a gradient in permeability is one aspect of the axial gradient, our conception of the relation between per-

1 Much of this work has been done by Mr. J. W. Mac Arthur and is not yet published. For some observations on algae see Child meability and the gradient must depend very largely upon the terms in which we define permeability. If permeability is dependent only on the physical condi- tion of the limiting surface and independent of chemical activity the axial gradients are manifestly something more than mere permeability gradients. But the pro- toplasmic limiting surface or membrane is certainly polyphasic in constitution, since it is alive and the seat of more or less chemical activity, and its permeability depends upon its living condition and changes when it dies. Moreover, susceptibility as determined by the higher concentrations and intensities of external agents depends rather upon the destruction or alteration of the limiting surface as a living membrane than upon the passage of the agent through the living membrane into the interior of the cell. The susceptibility gradients can be demonstrated not only by agents to which the protoplasmic surface is highly permeable, e.g., vital dyes, various anesthetics, but by those to which it is highly impermeable, such as mercury and copper salts, and by extremes of temperature and the negative con- dition lack of oxygen, which do not involve the action of any external chemical agent upon the surface.

Moreover, the phenomena of differential acclima- tion and recovery in growth and development indicate very clearly that the metabolic activity of protoplasm is a factor in susceptibility. With the advance of our knowledge it becomes increasingly evident that the factors concerned in the permeability of living proto- plasm are essentially those concerned in other aspects of life and that permeability is an expression of the physiological state of the plasma membranes. If we

admit this, differences in permeability are themselves to some extent indicators of differences in physiological state, but it still remains true that susceptibility is not simply a matter of the rate of penetration through the plasma membranes, but rather of the rate of killing or alteration of the membranes and superficial regions of protoplasm by an external agent. Different external agents may and undoubtedly do act chiefly or primarily upon different factors concerned in the maintenance of physiological state, but since these different factors are mutually associated in such maintenance, the general result as regards susceptibility, i.e., the general effect on the physiological state, may be and is the same for at least many different agents. In short susceptibility is within certain limits and in a general way an index of physiological state in protoplasm, and the axial gradients in susceptibility are therefore significant, par- ticularly when their existence is confirmed by other methods as indicating the existence of nonspecific or quantitative differences as the earliest distinguishable features of axiation.

The axial gradients have also been demonstrated in many forms as a differential in the rate and amount of reduction of potassium permanganate by protoplasm. It is a well-known fact that KMnO4 is reduced by pro- toplasm and the reduced salt appears on or in the protoplasm as a brown or blackish precipitate. All axiate forms examined, including numerous Protozoa, eggs, embryonic and larval stages or adults of the lower invertebrates, echinoderms and smaller arthropods and various algae among plants, show gradients in the rate of staining by permanganate corresponding to the gradients demonstrated by susceptibility and other methods. The precipitation of the reduced salt and the appearance of the brown color apparently begin on the external surface of the protoplasm, often within a few seconds after the organisms are brought into the solution, and the differences in rate of precipitation and staining at different levels of an axis are usually very marked. Penetration into the protoplasm usually occurs rather slowly, its rate depending somewhat on concentration, and it is certainly not to any large extent dependent on permeability of living membranes but rather on the killing of the protoplasm from the surface inward. If the reaction is allowed to continue to completion in excess of permanganate, the whole organism may become opaque black and no gradient is visible, but many small organisms thus stained, e.g., blastulae, hydroid planulae, small monosiphonous algae, can be made more or less transparent after such staining by hardening, clearing, and mounting, and in such cases the gradient in staining appears. Under these conditions the gradient represents a gradient in the total amount of reduction of permanganate of which the protoplasm is capable, and it is highly significant to find that the protoplasm of the apical region of a blastula or an alga axis, for example, is capable of reducing more perman- ganate than 'more basal levels. If the organism is first killed by some other agent, e.g., various histological fixing agents, heat, etc., reduction and staining are uniform, or in some cases slight traces of the gradients

are still present, for a time, but after a few days in alcohol they are completely absent in all cases examined. Since the chemical reaction concerned here is an oxidation-reduction the rate and amount of reduction of permanganate must be associated in some way with the oxidative activity of protoplasm, the regions of higher rate of oxidation showing a higher rate and greater total amount of reduction. We find that the permanganate gradients correspond with those demonstrated by other methods, i.e., the more active and more susceptible regions reduce permanganate more rapidly and in larger amount than the less active and less susceptible. This method is a very delicate one and I believe of consider- able value as a means of determining regional and axial differences in physiological state, particularly in small organisms.

In certain cases the indophenol reaction has been used to demonstrate the existence of axial gradients in blastulae and gastrulae. This reaction is an oxidation which is catalyzed by oxidizing enzymes and which results in the formation of indophenol in the form of a blue precipitate. In living starfish blastulae and gastrulae (Child, 191 5#) a distinct apico-basal color gradient has been observed by this means, the forma- tion of indophenol occurring most rapidly in the cells of the apical region. If the animals are killed by some other agent before the indophenol reaction, the blue color is much less marked and the gradient is absent.

Axial gradients in electric potential have been found to exist by Drs. Bellamy and Hyman in axiate animals so far as examined, the region of greatest physiological activity as indicated by other methods showing in general the highest electronegativity through the gal- vanometer. These data are not yet published. Obser- vations along these lines have been made by Mathews (1903) on hydroids, Waller (1903) on various organisms, and Hyde (1904) on certain vertebrate embryos, and both Mathews and Waller point out the probable relation between electric potential and physiological or metabolic activity. Morgan and Dimon (1904) in a study of electric potential in the earth\vorm found that in general the two ends were electronegative to middle regions and concluded that the potential differences were not related to physiological polarity. We know now, however, that the earthworm and other annelids develop very early a growing region of high physiological activity at the posterior end (Hyman, 1916; Child, 1917^) and that the body in later stages shows two gradients in opposite directions. The observations of Morgan and Dimon do not therefore conflict with those of others. More recently Tashiro (1917 and earlier papers) on the basis of his work on CO2 production in the nerve fiber has pointed out the probable relations between the electric phenomena and metabolic activity in nerve, and Hyman (1918) has also suggested that bioelectric phenomena in general are primarily due to differences in metabolic activity. While differences in potential undoubtedly may arise in organisms from other causes than differ- ences in metabolic or oxidation rate the facts in general indicate that such differences are at least very gener- ally associated with differences in metabolic rate. In any case the axial gradients in potential indicate the

existence of graded differences in physiological condition along the axes, and in the light of other data there can be little doubt that differences in rate of metabolism, or more particularly oxidation, play at least a very important part in determining the electrical gradients. The direct determination of metabolic activity in different body regions by means of oxygen consumption and CO2 production requires the separation of the regions concerned and therefore introduces various complicating factors. Only in the simpler organisms is it possible to maintain such separated pieces of the body in anything like a normal condition, and even here the operative procedure involves stimulation and may be followed by depression, and cell division and growth begin at the cut surfaces within a few hours after section. More- over, in organisms with localized and differentiated internal organs the axial gradients do not necessarily run in the same direction in all organs, and a piece of the body from a particular level may represent a low level of the body-wall gradient and a high level as regards certain internal organs, or vice versa. In Planar ia, for example, the mouth is near the middle of the body and the activity of the alimentary tract probably decreases from the mouth in both directions, but in the body wall the gradient is from the head pos- teriorly. If this is true, a piece of the Planaria body from levels near the head represents a high level of the body-wall gradient and a low level of the alimentary tract, while a piece from near the mouth represents

the reverse condition. Unless we can eliminate one or the other of these gradients the oxygen consumption and CO2 production of such pieces is not likely to give us any very definite information as regards either gradient. In this case, however, we can bring the alimentary tract into a more or less quiescent condition by starvation and then we find that a gradient in oxygen consumption and CO2 production does exist in the body wall, the high end being at the head with a second rise in the region of the posterior zooid. The data along this line are as yet mostly unpublished, but one paper on CO2 production in Planaria has already appeared (Robbins and Child, 1920) and work on oxygen con- sumption has been done by Dr. Hyman with similar results. But the relative amount of alimentary tract tissue in pieces of a given weight taken near the head is less than in pieces taken near the mouth, so that even in animals which have been starved for some time the difference in total oxygen consumption or total CO2 production of pieces from different regions is not strictly comparable with the regional differences in suscepti- bility of the body wall. Where it has been possible thus far to use these methods we have found that the axial gradients demonstrated by other methods are also gradients in oxygen consumption and CO2 production.

By one method or another or by several different methods the existence of physiological gradients in various axiate organs or parts of many organisms has been demonstrated, e.g., in various reproductive axes and in the " hairs " of algae (Child, 1916^, e, 191 ja, iqige), in the larger, slow moving flagellum of Noctiluca, in the tentacles of Hydrozoa (Child and Hyman, 1919; Child 19196) , in the plate rows of ctenophores (Child, 191 yc), the growing arms of echinoderm larvae, the branchiae and sensory tentacles of various annelids, the growing tail of the ascidian and the amphibian tad- pole, etc. Dr. Hyman has found that the embryonic heart of the chick and of the fish represents a suscepti- bility gradient with the high region at the sinus end. But the most extensive work on the metabolic gradient of any organ is that of Alvarez and his assistants on the vertebrate alimentary tract.1 They have found in the small intestine a gradient in irritability, latent period, tone rhythm, conduction, susceptibility to various drugs, and gradients in at least some of these conditions in the wall of the stomach and colon. Tashiro (1917 and earlier papers) has found a gradient in CO2 production in cer- tain nerves, the direction of functional conduction being down the gradient and in certain of these nerves a susceptibility gradient has been observed (Child, 19140). In an investigation of the respiration of ground nervous tissue C. G. MacArthur and Jones (1917) have found differences in rate of respiration in different parts of the central nervous system which indicate the existence of a gradient in rate of respiration. The rate of respira- tion is highest in the cerebrum and decreases in the various parts in the following order: cerebellum, mid- brain, medulla, corpus callosum, spinal cord nerve. The authors find that gray matter consumes about twice as

much oxygen and produces about one and one-half times as much CO2 as white matter, and some of the differences in respiratory rate in different parts of the nervous system are doubtless due to differences in pro- portion of white and gray matter. For example, the relatively low rate of the corpus callosum is undoubtedly associated with the fact that it consists of white matter, nerve fibers, rather than cells. But the differences in respiratory rate between cerebrum and cerebellum and between midbrain, medulla, and spinal cord are scarcely to be accounted for in this way. These differences con- stitute highly significant evidence for the existence of an axial gradient in rate of respiration in the central nervous system. More recently MacArthur and Doisy (1919) have demonstrated the existence of a gradient in chemical differentiation in the brain.

All the evidence is in agreement as regards the existence of these physiological axial gradients. Thus far every living physiological axis examined has given evidence of the existence, at least in the earlier develop- mental stages of such a gradient, and in many cases the experimental methods show the presence of a gradient wrhere structural, or other directly visible indication of its presence are absent. In whatever terms we may finally interpret these gradients, there can be no doubt concerning their existence. They are physiological facts, and their significance for localization, differen- tiation, and functional relation is already demonstrated.

It is a familiar fact that physiological axes persist through certain agamic reproductive processes and are therefore inherited by the individuals resulting from such reproduction, but in at least many eggs the axiate pattern apparently arises de noio during the growth of the egg. Granting that axiate pattern in its simplest form is a gradient pattern, as pointed out, we have at present no grounds for believing that such pattern is inherent in protoplasm, or that it can arise in protoplasm apart from the action of environmental factors. Appar- ently a differential exposure of the protoplasm to some environmental factor or factors which affect its rate of activity, i.e., which are primarily quantitative rather than specific or qualitative in their action, is necessary for the origination and establishment of an axial gradient in protoplasm. Moreover, the quantitative nonspecific character of . the gradient pattern suggests that its origin is to be sought in the quantitative rather than in the qualitative or specific aspects of the relations between protoplasm and environment. In other words the gradients must arise through the differential action of environmental factors which affect primarily the rate of general protoplasmic activity.

As regards the plants, Winkler ( 19006) and Kniep (1907) showed that in certain algae, e.g., various species of Fucaceae differential illumination of the two sides of the egg determines the polarity of the plant developing from the egg, though in the absence of light polarity appears and germination occurs, but more slowly. Stahl (1885) showed that in the spore of Equisetum polarity is determined in the same way. Winkler (1900^) also showed that polarity in the alga Bryopsis could be determined by light. In various bilaterally symmetrical plants, such as liverworts, light determines the dorsiventrality and some algae develop a radially symmetrical thallus wrhen the illumination is equal on all sides of the polar axis and are bilateral when the illumination is from one side. In the spermatophytes conditions which determine egg polarity are undoubt- edly intraorganismic. The embryo sac shows a definite polarity with respect to surrounding parts and the ovum is attached to one end of the sac. The unattached end of the ovum becomes the apical, the attached end the basal pole of the embryo. What particular factors are concerned in this case is not known.