Individuality in Organisms

correspond to these different types of hcad^ and are undoubtedly the fundamental factors in determining general form and localization of the parts in the head"! Figs. 54-56. — Different results of reconstitution in Planaria dorotocephala: Fig, 54^, teratophthalmic animal; Fig. 54^, dilTercnt forms of eyes in teratophthalmic animals; Fig. 55.4, B, teratomoqihic forms; Fig. 56^, B, anophthalmia forms. As a matter of fact, these different forms are a more or less arbitrary grouping of what is actually a graded scries of forms from the normal head at one extreme to the

^ See Child and McKie, "The Central Nervous System in Teratophthalmic and Teratomorphic Forms of Planaria dorolocep/hila," Biol. Bull., XXII, 191 1. headless form at the other. I have determined experimentally that these different forms represent different degrees of retardation or inhibition of the process of head formation. Their formation can be controlled experimentally in a great variety of ways. For example, the percentage of pieces producing heads, which we may call the headfrequency, is less in shorter than in longer pieces, in pieces from more basal than in those from more apical levels of the body ; less in pieces from young than in pieces from old animals, in pieces from starved than in pieces from well-fed animals, in pieces which are kept quiet than in those forced to move about.

The effect on head-frequency of substances which decrease metabolic rate, such as dilute solutions of cyanides and narcotics, is of great interest, for it is definite and modifiable experimentally, but not uniform. In series of pieces of equal length, a, 5, c, Fig. 57, taken from animals of the same size and as nearly as possible the same physiological condition, the head-frequency under natural conditions is highest in the a-pieces which represent the most

Fig. 57. — Outline of Planaria dorotocephala, indicating regions, a, b, c, from which pieces are taken. the Z>-pieces it is lower, and in the c-pieccs lowest of all. If such a series of pieces is placed for a few hours after cutting in a low concentration of cyanide, alcohol, etc., the head-frequency in the a-pieces is considerably lower than in water, that in the 6-pieces slightly lower or about the same as in water, while that of the c-pieces is higher than in water. This result is characteristic, but the actual percentages can be altered by differences in concentration of the reagents, temperatures, and many other factors.

Although at first glance these results appear hopelessly confusing, they depend upon a very simple relation between that region of the piece which gives rise to the head and other parts. In an isolated piece of the planarian body (Fig. 58) the head arises from the cells of the region x, which are more directly affected by the wound and undergo rapid dedifferentiation and rejuvenescence and so attain a higher metabohc rate than cells farther away from the cut surface and begin soon after section to divide and grow rapidly. If these cells give rise to a head, the region y undergoes more or less transformation to form the body of the new individual. I have found that the head-frequency varies directly with the metabohc rate in x, the head-forming region, and inversely with the metabolic rate in the region ;•. This relation may be stated in the formula, headfrequency = ?^ . This means that the higher the mcta

Fig. 58. — Diagrammatic outline of a piece of Planaria to illustrate relations of new apical region, .v, new basal region, :;, and old body region, y. to a head, and vice versa, and it also means that the higher the metabolic rate in the region y, the less likely the piece is to give rise to a head. If this relation is altered by an increase of rate x relatively to rate y, headfrequency is increased; if by an increase in rate y relatively to rate x^ head-frequency is decreased. On this basis all the experimental effects of different physiological and external conditions on head-formation can be readily accounted for, and it has even been possible in many cases to predict the results of various experiments.

Some of the facts on which this conclusion is based are as follows: By means of the susceptibility method I have demonstrated that the act of section always increases metabolic rate, particularly in the part basal to the cut. This condition of stimulation continues in the pieces for several hours after cutting and only gradually disappears.^ The more basal the level of the piece in the original body, the more its metabolic rate is increased by section. In the cases of pieces a, 6, c in Fig. 57 the metabolic rate during the first few hours after section is higher in h than in a and higher in c than in &, although before section the rate decreased from a to c. This difference in stimulation of pieces from different levels results from the different degrees of subordination. The region c is subordinate to all more apical regions and is much more dependent upon impulses coming from these regions than is the region a, which is subordinate only to the head. When the chief paths of conduction in the nervous system are cut they

' Child, "Studies on the Dynamics of Morphogenesis. VII," Jour, of Exper. ZooL, XVI, 1914. are stimulated ; consequently the more basal the levc! of a piece the more its rate is increased by section. I have also found that the shorter a piece, the higher the metabohc rate after section. Long pieces are stimulated but Httle, except at the ends, chiefly the apical, but in short pieces the rate increases greatly. Another simple experiment' shows that under ordinary conditions it is determined within three to six hours after section whether or not a head will develop on a piece. This is during the period of stimulation of the piece, and when we compare head-frequencies and metabolic rates during the period of stimulation following section, we see that the higher the metabolic rate in the piece as a whole, i.e., the region y, Fig. 58, the less likely a head is to develop, and vice versa. The headfrequency is lower in more basal pieces such as c than in more apical pieces like a, and in shorter than in longer pieces, because the metabolic rate in the region y is higher at the time of determination of the course of development. Pieces from young or starved animals also have a higher metabolic rate^ and a lower head-frequency than similar pieces from old or well-fed animals.

These facts may seem to involve a paradox, but their interpretation is actually simple. The two regions X and y (Fig. 58) of the piece behave differently after section. The cells of x are so extremely affected by the presence of the wound and the altered conditions that they rapidly dedifferentiate and begin to divide and grow, and so approach or attain an embryonic condition. The cells of y^ however, merely undergo a temporary increase in metabolic rate. The region a: is a small groups of cells undergoing dedifferentiation, while y represents a considerable portion of a fully developed individual with established relations of parts and specialized nerves which are much more efficient than embr^^onic protoplasm as conducting paths. The region X originally has a higher metabolic rate than y, because it represents a more apical level in the gradient, and its rate rises still farther as it begins to dedifferentiate. The facts of experiment indicate that in order to produce a new head rate x must not merely be higher but much higher that rate y. The relation between x and y is evidently this: if rate x is sufficiently above rate y, x develops independently of y into a head and dominates y, while otherwise y dominates a: to a greater or less extent and so retards or inhibits head-formation, and the various forms between normal and headless condition are produced.

This relation ^^^ can be altered in various way-s: by means of dilute narcotics it is possible according to the method of use either to decrease the stimulation in y resulting from section or to delay the reaction in x until after the increased rate in y has largely or wholly disappeared, or finally the relation may be altered by the more rapid and more complete acclimation of the young cells at X as compared with the older cells of y (see pp. 51, 52). In pieces such as c. Fig. 57, where under ordinary conditions x and y are, so to speak, evenly matched in the struggle for dominance and the headfrequency is low, all these methods increase the headfrequency, because the relative increase in rate of x as

compared with y overbalances the absolute decrease in rate produced by the narcotic. In pieces like a, where y is only slightly stimulated by section and rate x is so much higher than rate y that the head-frequency is very high, the effect of narcotics is to decrease headfrequency, because in such cases dominance is not reversed and only the direct inhibiting effect of the narcotic on the region x appears. Head-frequency may be increased in all pieces by inducing them to move about.' The apical end, of course, precedes in such movement, the cells of the region X are subjected to more excitation than in a piece which is not moving, and the higher metabolic rate of x results in increased head-frequency.

In Planaria maculata and certain other species the degree of subordination of basal regions of the body is not as great as in P. dorotocephala; consequently the increase in metabolic rate after section in pieces from this region is less than in P. dorotocephala, and in these species the head-frequency of such pieces is almost or quite as great as that in more apical pieces. Various other differences in the reconstitutional process in different species of planarians only serve to confirm the conclusions reached in the case of P. dorotocephala.

These and many other facts have forced me to the conclusion that the head wliich appears in the reconstitution of a piece is not physiologically a part of the piece and is not formed by the piece, but develops, so to speak, in spite of it. Only when the metabolic rate of the cells at x is high enough to make them essentially ' Child, "Experimental Control of Morphogenesis in the Regulation of Planaria,^' Biol. Bull., XX, 191 1. independent of y do they begin the development of a new individual by the formation of a head. The formation of a head at the end of a piece is then exactly the same process as the transformation of short pieces into heads when no other part of the body is present (pp. 96-101). The new head arises independently of other parts and dominates them. The influence of other parts on head-formation is merely inhibitory or negative, while the influence of head -formation on other parts is determinative or positive. The process of development of the cephalic ganglia in the head formed on a piece also indicates the independence of the head. The ganglia arise in the new tissue independently of the parts of the nervous system in the old tissue of the piece and become connected with these parts only secondarily.' This fact suggests that head-formation actually depends upon the establishment of a melabolic gradient in the region x with its apical region near the free end and decreasing in rate toward y. If this occurs, a head forms, but if rate y is high enough in relation to rate x, headformation is inhibited or retarded by the interference between two gradients in opposite directions. Inhibition or retardation of head-formation consists then in the interference of one metabolic gradient with another in the opposite direction or the obliteration of the one by the other.

The new basal end of the piece develops from a grou[) of cells 2, Fig. 58, which react to the wound at the basal end by more or less dedifferentiation and growth. This ' S. Flexner, "The Regeneration of the Nervous System of Pianaria lorva," etc., Jour, of MorphoL, XIV, 1898; Child and McKie, "The Centjal Nervous System in Teratophthalmic and Teratomorphic Forms of Pianaria dorotocephala,'" Biol. Bull., XXII, 191 1. reaction is less rapid than that of x (see Fig. 31), because they represent a lower level in the gradient and their relation to the region y is different from that of x. The rate of development and completeness of the new basal end varies directly with the metabolic rate in y; any conditions which decrease the metabolic rate in v decrease the development of the basal end, and vice versa. We may say then that tail-frequency = "]-^-^ but that under the usual conditions, when a gradient is already present in the piece, rate z is so low that it becomes negligible, and the formula becomes tailfrequency = rate y. This holds true as long as a new zooid does not arise in this basal region. If a new zooid does arise there in consequence of physiological isolation, as is often the case in headless pieces, then the lower the rate in y, the more rapid the development of this posterior zooid. In headless pieces the large size of the posterior outgrowth (cf. Figs. 52, 53, with Fig. 50, p. 103) is due to the fact that this region is not physiologically the basal end of the individual but a second individual.' The development of the basal region is then dependent upon the presence and influence of more apical regions, while the development of the head occurs independently of other parts, so far as it is not inhibited by them. The relation between the major axial gradient and these differences of behavior in different regions is evident. The process of reconstitution of a new individual from a headless piece in Planaria is a process of development beginning at two different levels, first, at the apical end of the piece with the formation of a

^ Child, "Studies on the Dynamics of Morphogenesis. TIT," Jour, of Exper. Zool., XI, igii. new head, and, secondly, at the basal end with the formation of a new tail. The new apical region as the region of highest metabolic rate determines the establishment of a new major axial gradient, which has the same direction as the original gradient but possesses a higher rate, and in consequence of these changes the parts of the piece below the new apical region undergo more or less structural change into parts characteristic of more apical levels, until sooner or later a stable condition of the gradient is attained, and this determines the completion of reconstruction.

It can scarcely be doubted that the process of reconstitution of pieces into new individuals is fundamentally the same in all animals, though it may differ widely in details, with the kind and physiological condition of the individual or piece and the nature of the external conditions under which reconstitution occurs. Moreover, it is essentially the same process as reconstitution in plants, in that it consists in the development of a new individual beginning with the apical end. The chief difference is that in animals the development of the new individual is usually closely associated with the cut surface or surfaces, while in plants the reaction of the cells at the cut surface usually does not at once cover it with more or less embryonic rapidly growing cells, as it does in animals, and, since the plant is usually a composite individual, other apical regions already present become dominant, or new apical regions arise in other parts before a new apical region develops at the cut surface. In some cases, where only a small part of the apical region is removed, a new growing tip develops from the cut surface, and in such cases the formation of the new grow-

ing tip is, I believe, essentially the same process as the formation of the new head in Planaria. In cases where wound callus develops, new growing tips may arise in that. In the formation of a new growing tip in callus tissues (pp. 85-86) and its later connection with other parts of the plant we have again a process very simihir to the formation of a head in a piece of Planaria, and the development under its dominance of other parts, so far as they are not already present. In both cases the new apical region is not determined by other parts but develops independently of them, and its later relations to them are determined by its own dominance.

The development of double or biaxial apical regions from short pieces has been discussed above (pp. 98, 99). In some cases biaxial basal regions instead of apical regions arise from pieces. Pieces of the stems of certain hydroids sometimes produce stolons at both ends, biaxial tails have been p^^ 59 — Experiobserved in short pieces of Planaria mentally determined by Morgan, and I have been able to reconstitution of produce them experimentally m some ^.^^^ ^^ pianaria. cases (Fig. 59) by altering the relations of metabolic rate between the regions x and y (Fig. 58) with the aid of narcotics. In the earthworm and related forms various investigators have observed the development of tails at both ends of pieces from the more basal regions of the body. My own experiments indicate that when the development of the new tissue at a cut end of

a piece is completely dominated by the piece it gives rise to a basal structure. Such dominance means simply that the old tissue has a high enough metabolic rate to determine the direction of the gradient in the new tissue. In Planaria the development of tails at both ends of a short piece is apparently due simply to the fact that the metabolic rate in the piece is high enough so that the new tissue does not become dominant at either end but develops under the control of the old tissue. Dr. Hyman has found that the conditions determining the formation of double tails in Lumhriculus seem to be essentially the same as in Planaria, though the factors which produce them are somewhat different. She has been able to determine experimentally to some extent the production of heads instead of tails in such pieces by methods similar to those which I have employed for altering head-frequency in Planaria. She has also observed the development of structures intermediate between head and tail, or rather inhibited, rudimentary cephalic ends, in which certain caudal characteristics appear later. These are apparently cases in which the new tissue was at first to some extent independent but later became subordinated to the old.^

The absence of any outgrowth at the apical end of a piece, as in the headless forms of Planaria (Figs. 52, 53), occurs when head-formation is completely inhibited, but the degree of dominance is not sufficient to determine development as a tail. In such cases local conditions at the cut apparently determine the result and the wound simply heals. In some other forms the wound ^ I am indebted to Dr. Hyman for permission to use these unpublished data.

reaction involves more growth than in Planaria, and in such cases considerable outgrowths may sometimes arise which are neither heads nor tails, but cell masses of indeterminate character which gradually differentiate in relation to adjoining parts and may finally show both apical and basal characteristics. In many of the flatworms and various other forms only an apical cut surface above a certain level of the body gives rise to a head, while tails may arise from cut surfaces at any level basal to the head of the parent body. In some of these cases the level where headformation ceases lies a considerable distance from .the cephalic ganglia, while in other cases head-formation does not occur when the cephalic ganglia are removed; but when parts of the head are removed leaving a portion of the cephalic ganglia intact — sometimes half or more, sometimes only a small part, is necessary — such parts develop again. In the headless pieces there may be more or less outgrowth at the apical end of the piece, but it is indeterminate in character. Some authors have maintained that in such cases the cephalic ganglia or the more apical regions of the longitudinal nerve cords exercise a specific formative influence of some sort and so determine the development of a new head, but there is no real evidence in favor of this \'icw. Probably the head fails to develop in such cases either because the cells reacting to the wound do not attain a high enough metabolic rate to become independent of other parts and their development into a head is therefore inhibited, as in the headless pieces of Planar ia. or because these cells do not dedifferentiate to a sufficient extent to be capable of giving rise to a new cephalic

ganglion and so to a head, while they may still be able under the dominance of other parts to produce basal regions of the body. The development of the apical region or of the apical part of the central nervous system, which in all except the lowest animals is the primary and dominant part of the apical region, is a self-determining process, independent of other parts, while the development of other parts is determined by their relations to dominant regions. It is highly probable therefore that a more complete loss of differentiation is necessary as a condition for head-formation than for the development of other parts. As a matter of fact, we find that as the capacity for reconstitution becomes limited by increasing differentiation the capacity for head-formation disappears first of all. Many animals in which reconstitution of new heads does not occur are still able to reproduce all subordinate parts, and with further limitation it is the more subordinate parts, such as legs and other appendages or caudal regions, which the body is capable of reproducing.

This limitation is more or less progressive from lower to higher forms, until in the higher vertebrates the capacity for reconstitution under any known conditions is limited practically to tissue regeneration. The primary limiting factor is unquestionably the increasing physiological stability of the protoplasmic substratum, in consequence of which the capacity for dedifferentiation and rejuvenescence, at least under ordinary conditions, is more and more narrowly limited.^

The embryonic stages of different animals differ widely as regards their capacity for reconstitution. In the sea-urchin and starfish isolated cells or groups ol cells of the developing embryo down to a certain limit may give rise to complete larvae of small size, while in other forms, such as the annelids and mollusks, isolated parts of the embryo show little or no reconstitutional change, but remain alive for a time and continue to differentiate as they do when they remain as parts of an intact embryo. From the failure of the isolated parts to undergo reconstitution the conclusion has been drawn that they are independent of each other in the intact embryo, and that development -in these organisms is a sort of mosaic made up of independent parts with some sort of pre-established harmony between them. If this view is correct, there is no relation of dominance and subordination in these stages of development. The failure of isolated parts to undergo reconstitution does not, however, demonstrate the absence of dominance but merely the ineffectiyeness of isolation. The absence or limitation of embryonic reconstitution in certain forms is apparently due, like the increasing limitation of reconstitutional capacity in higher animals, to the higher specialization of the parts of the egg and embryo in these forms. There is good reason to believe that in such eggs the condition in embryonic stages is the result of differentiation dependent upon dominance and subordination of parts in the earlier life of the egg, and that specialization has gone beyond the stage where it can be greatly altered by isolation. Development proceeds in isolated parts as far as it has been determined by past relations with other parts or as

far as nutritive or other conditions permit, and then ceases. There can be Uttle doubt that relations of dominance and subordination exist during embryonic stages, and that these are factors in determining what occurs in later stages. According to this view, the difference between these eggs and those in which a high degree of embryonic reconstitution occurs is primarily a difference in the stability or fixity of the effects of previously established metabolic gradients. At the one extreme are eggs in which axial differences at the beginning of embryonic development are probably largely or wholly differences in metabolic rate, at the other, those in which specialization and differentiation of parts have gone far beyond this condition. The egg, in short, is an individual, and some eggs are more highly specialized individuals than others.

The proportional relations of parts in reconstitution, of which much has been made by Driesch, Morgan, and others, are obviously, so far as they exist, dependent upon metabolic relations between the parts. On a short piece of Planaria, for example, a smaller head usually develops than on a long piece. This fact has often been regarded as in some way associated with the fact that the shorter piece will produce a smaller animal than the longer and that the size of the new head foreshadows the size of the animal. As a matter of fact, the size of the head formed by pieces of the same size may differ widely in different cases and can be controlled experimentally to a very large extent by controUing metabolic conditions. The higher the metabolic rate in the region x, Fig. 58, in relation to that of the region y, the larger the head, and vice versa. The size of tJie

head in relation to other parts is determined primarily by its ability to grow at their expense. In a shorter piece there is less material available for such growth than in a longer piece, consequently a smaller head develops. Essentially the same relation exists as regards other parts. Where an excess of nutritive material is available the relation is not necessarily very different, for each part uses nutrition instead of the substance of other parts according to its metabolic activity, i.e., according to its position in the axial gradients, so that in this case also the chief factors in determining the proportions of parts characteristic of each form are the metabolic relations between them. In the early stages of development in nature the simple quantitative gradation in size from the apical toward the basal region appears, but as specialization occurs and the differences in metabolic rate at different levels bring about changes in metabolic character the size relations must of course become more complex.

The return or approach to the characteristic form of the species which very commonly takes place in the reconstitution of pieces has been regarded by JMorgan and others as largely a matter of the physical rearrangement of the substance of the piece. That changes in shape may be brought about in soft-bodied forms like the flatworms by mechanical conditions connected with motor and other functional activities of the animals, I have shown.' Wherever such factors play a

part in determining the characteristic shape of the animal they undoubtedly play a part in determining the approach to this shape in pieces undergoing reconstitution, but in cases where they are not primarily concerned the metabolic relations are unquestionably the primary factors in determining shape and proportions of the whole and parts. In most adult animals and embryonic stages which are capable of any considerable degree of reconstitutional reproduction, a limit of size of isolated pieces seems to exist below which reconstitution becomes incomplete or fails to occur. In Planaria, for example, with decrease in size of piece head-frequency falls to zero, but with still further decrease in size head-formation begins to occur again and head-frequency rises. These changes are simply due to changes in the relation JJ^ (see pp. 109- 10). With decreasing size of the piece, y is more and more highly stimulated by section until in pieces below a certain size heads do not develop at all, but when the piece becomes very small y practically disappears, for the whole piece becomes involved in the direct wound reaction and so corresponds to the region x or such a region in relation to both cut ends. In such pieces there is nothing to inhibit or retard headformation except the simultaneous development of a head at the opposite end (see pp. 98-101), and in such cases the effect is mutual and results merely in retardation.

Here then the completeness or incompleteness of reconstitution in relation to size of piece is wholly a matter of quantitative metabolic relations. There is no minimal size of piece which represents the '^organization" of the species reduced to its lowest terms. The minimal size can be altered widely even now by controlling conditions, and I have no doubt that if we arc ever able to isolate single cells and to provide proper nutritive and other conditions for them we shall fmd that in many of the lower animals such cells are capable of giving rise to new individuals, as they undoubtedly are in many plants.