Physiological Foundations of Behavior

Little is known as yet concerning electricity as a factor in determining axiation in animals. As already noted, Ingvar (’20) has observed the determination of apparent polarity in tissue culture by the electric current and Lund (21 a, ¢) has reported determination of polarity in pieces of hydroid stem by the same means. The suggestions advanced in the preceding chapter and elsewhere (Child, ’21 a, Chap. XI) concerning the réle of bioelectric currents in determining neuron polarity are as yet without experimental basis, though it seems difficult or impossible to account for the facts in any other way than through some sort of differential action of the bioelectric currents on the cells. I am inclined to believe that future investigation will show that electric factors in the proper intensity and relation are highly effective in determining physiological gradients and so axiation. As regards the manner in which the electric current acts nothing definite is known, but, as suggested above (pp. 111-112), it seems to be possible to account for the facts by changes in electrical polarization of cells or cell masses and the effect of such change upon metabolic rate and physiological condition in general.

| Gravity as a possible factor in determining specific localization and differentiation has been considered in an earlier chapter (pp. 54-55). Here we are concerned with its possible action in originating and establishing physiological gradients. In various plants radial and bilateral symmetry show essentially the same relation to gravity as in other cases to light. Such plants or parts are radially symmetrical when their longitudinal axes coincide with the direction of gravity and dorsoventral when in other positions. It has been known, for example, that in pieces of stems of various plants in horizontal position, roots tend to appear on the lower and shoots on the upper side, and it has been experimentally demonstrated that the essential factor in such cases is gravity. Within recent years Loeb (’17 a, b, 719, 20 and other papers) has repeated many of the older experiments and added some further data, besides attempting an interpretation of the action of gravity in terms of transport of chemical substances, hormones, or formative substances. In these cases the action of gravity does not obliterate the preéxisting polarity but merely modifies it to some extent. Thus far no case is known in which gravity determines polarity in the egg, the spore or the growing tip of the plant as light does in various cases. In general, gravity is far less important than light in determining axial relations and forms in plants. Loeb (’92) has described the determination of polarity by gravity in an animal, the hydroid Antennularia. Up to the present this case remains unique. Morgan (’01 a) and Stevens (02, ’10) have shown that gravity is certainly not essential in this case, since polarity may be determined by other factors.

As regards the way in which gravity produces its effects we are even more in the dark than as regards the action of light and electricity. Transport of specific substances has often been postulated since Sachs, and Loeb in his recent papers (17 a, b, "19, ’20) attempted to interpret the facts in essentially these terms. ‘This transportative conception fails, however, to account satisfactorily for many facts and interpretation must apparently be sought along other lines. It is perhaps permissible to call attention to a suggestion recently cited by Harvey (20, p. 367) from a personal communication of R. 8. Lillie. This suggestion is essentially that the distribution of sap by gravity may determine differences in electrical conductivity, and so differences in bioelectric currents in different regions. It seems probable at any rate that the action of gravity will prove to be

_ primarily quantitative rather than specific as regards different regions of the plants. As regards polarity its action is not fundamentally different from that of light, which is apparently primarily quantitative. In other words, the action of gravity in determining polarity or symmetry, so far as such action occurs, is probably essentially like that of other external factors, a determination in some _ way of a quantitative gradient in physiological condition.

The appearance of new axes in many cases of regulatory development is very evidently associated with differential action of an external factor, even though the factor of the environmental complex which is chiefly or primarily concerned has not been determined. New gradients determined by differential exposure. — Pieces 10 mm. or more in length from the stem of the hydroid Tubularia (Fig. 91) usually develop a hydranth first at the apical, and slightly later at the basal end (Fig. 92) when both ends are equally exposed. That the new polarity at the basal end is determined by the action of some external factor is clearly shown by two facts: first, it was not originally present, and second, if the basal end is closed by wax, paraffin, etc., or even buried in the sand, the new polarity fails to develop. In general, the shorter the piece of stem isolated the less clearly does the original polarity appear and the more rapidly do new polarities arise in response to external differentials. As I have pointed out elswhere (Child, ’15 ¢, pp. 96-102) this general disappearance of the original polarity with decrease in length of the piece is to be expected if polarity is a physiological gradient, because the shorter the piece the more nearly are its two ends physiologically alike. But if polarity is a molecular phenomenon it is difficult to account for its disappearance with decreasing length of piece. Such short pieces of the stems of Tubularia, Corymorpha, and various other hydroids may give rise to single (Figs. 98, 94) or to biaxial forms (Figs. 95-98), according as a single or a double gradient is determined. In the biaxial forms one of the axes is obviously new and determined in relation to the cut surface at that end. In general, the shorter the piece the more frequent biaxiality, and this fact shows very clearly that as the length of piece decreases the external factor becomes more effective than the original polarity.

Similar biaxial forms appear in short pieces of planarians (Figs. 99-101) and it is possible to increase their frequency experimentally by partially obliterating the original gradient by means of anesthetics or other inhibiting agents at the time of, and for a certain period after section (Child, 715 c, p. 149). In all these and in many other cases Figs. 91-98.—Reconstitution in Tubularia: Fig. 91, young, unbranched individual; Fig. 92, usual result of reconstitution in a long piece of stem with origin of a new gradient at aboral (lower) end; Figs. 93, 94, single apical structures from short pieces: Figs. 95-98, biaxial apical structures from short pieces (from Child, ’15 ¢).

the direct exposure to the medium of the cells at the cut ends, permitting a more rapid respiratory exchange than in other regions where the body wall is intact, and the growth and cell division following the wound are undoubtedly factors in establishing the new axes. These new axes are in all cases represented by physiological gradients and these gradients apparently result simply from the gradation ~ In exposure and its effects from the cut surface inward. Hypotheses of reversal in orientation of cells or molecules are quite superfluous, and there is no reason to believe that exposure at a cut surface can bring about such changes.

Differential exposure after experimental obliteration of original gradients. — In Corymorpha, a hydroid much like Tubularia, except that most of the stem is naked, the original polarity of stem pieces (Figs. 102-105) or even single or double hydranths (Figs. 106-109) can be obliterated through differential inhibition with various agents and then on removal of the inhibiting agent a new polarity arises, _ determined by the differential between the free surface and that in contact (Figs. 104, 105, 108, 109). Since the new polarity is indicated

by a gradient involving differences in rate of respiration, the rate being highest on the free surface, it seems probable that the gradient is not some peculiar reaction to contact and its absence, but rather a matter of the difference in oxygen supply and perhaps in rate of CO, removal between free and attached surfaces. Similarly in the development of sponges from dissociated cells (H. V. Wilson, ’07, "11 a) the polarity of the new individual is evidently determined by the differential between the free surface and the surface in contact with the substratum.

Determination of new gradients by localization of growth. — The case of the sea-anemone Harenactis, in which it is possible to localize new axes by means of injury and the resulting growth (Child, - 709,10 b, 715 c, pp. 146-9), requires mention here. When short cylindrical pieces are cut from the body of this animal as indicated in Fig. 110, a, b) and most of the mesenteries removed from inside them, they close, as indicated in Figure 111, so that apical and basal cut surfaces of the body wall unite about the whole circumference and more or less new tissue arises along the line of union. This method of closure gives rise to “ rings” or doughnut-shaped forms and the line of union

usually comes to lie somewhere on the outer surface in consequence of a peculiar revolution of the ring upon itself, like the movement in a vortex ring, presumably an attempt at orientation. From both sides of the line of union tentacles and tentacle-groups arise, some bilateral, others radial and still others asymmetric (Figs. 112, 113). The tentacle-groups tend to appear in regions where the two cut surfaces did not fit closely together and where consequently considerable growth

Figs. 102-109.—Experimental determination of new polarity in pieces of stem of Corymorpha: Fig. 102, a piece of stem after isolation; Fig. 103, reduction and “melting down” in dilute alcohol; Fig. 104, appearance of new hydranth on upper side of mass after return to sea water. This polarity is at right angles to original axis; Fig. 105, later stage of new individual, with dotted lines indicating earlier outline of piece as preserved by thin perisarcal secretion; Fig. 106, a short piece which has developed biaxial hydranths and is therefore bipolar; Fig. 107, reduction of biaxial piece in dilute alcohol involving disintegration of tentacles and dedifferentiation of hydranth bodies; Fig. 108, appearance of new hydranth on upper surface of piece after return to water, the new polarity being at right angles to the former bipolar axes; Fig. 109, later stage of new individual, with dotted lines indicating earlier outline of piece as preserved by perisarcal secretion (from Child, ’15 ¢).

of new tissue occurs, giving rise both to body wall and mesenteries. Where the cut edges were closely apposed and little new growth has occurred, few or no tentacles arise. These facts led to the experiment of injuring locally a portion of the cut body wall by repeated snipping and complete removal of mesenteries. Such regions were found to give rise to larger and more nearly normal groups of tentacles than regions where less growth occurred. Figure 114 shows a case in which a normal axis arose from such an injured region. In this case the new axial gradient originates in the gradation in rate of

-metabolism, growth, etc., which has its high region in the area of injury and greatest development of new tissue. That the kind of symmetry which arises may differ widely in different cases according to the shape and extent of the regions of more active growth is indicated by the Fias. 110-114.—Localization of new axis in sea anemone, Harenactis, by localized injury and growth: Fig. 110, outline of animal indicating pieces, a, b, used; Fig. 111, closure of pieces, oral end uniting with aboral end about whole circumference to form a “ring’’; Figs. 112, 113, Tentacle groups developing along the line of union of oral and aboral cut ends in such rings; Fig. 114, a new normal axis localized by localized injury with resulting growth in a certain region of circumference.

various symmetries of the tentacle-groups in Figures 112 and 113. In connection with this case of the localization of a new axis by determination of more rapid growth through local injury, a statement of Harrison’s concerning the amphibian leg primordium is of interest. He says: “The limb rudiment may be thus regarded, not as a definite circumscribed area like a stone in a mosaic, but as a center of differentiation in which the intensity of the process diminishes as the distance from the center increases, until it passes away into an indifferent region. Many other systems, such as the nose, ear, hypophysis, gills, seem to have the same indefinite boundaries, which may even overlap each other”’ (Harrison, ’18, p. 456). It is perfectly clear from this statement that Harrison regards the primordia of these various organs as physiological gradients in a more or less specialized cellular region of the embryo, though he does not believe that such gradients are the fundamental factors of physiological axes. In a more recent paper (Harrison, ’21) he postulates a molecular symmetry and asymmetry as a basis for the symmetry relations in the amphibian leg (see pp.

Adventitious buds. — There are many other cases in which a new axial gradient arises as the physiological gradient from<a localized ne 4 ( Fies. 115, 116.—Origin of adventitious buds from epidermal cells of Begonia leaf: dicates gradient in protoplasmic structure from uniformly granular, deeply staining cytoplasm in more deeply shaded cells, to more or less vacuolated, less deeply staining cytoplasm in less deeply shaded and unshaded cells. region of growth to adjoining regions. The formation of adventitious buds in plants is a case of this kind. In Begonia, for example, buds arise from the specialized epidermal cells of the isolated leaf, or in some cases without isolation (Regel, ’76). In these buds the new gradient is directly visible as a gradient in cell size and rate of division and protoplasmic content (Figs. 115, 116). New axes appear in the same way in many other plants. While certain regions of the leaf or other parts usually show more. or less predisposition to adventitious bud-formation as compared with others, the localization of each individual bud in a particular epidermal cell or cell group must be due to a slight fortuitous difference between this cell or cell group and

others about it. Such difference determines that this cell or group shall react more quickly than those about it to the altered conditions and this reaction originates the new gradient and the new plant axis. When we say that the difference is fortuitous, we mean merely that we do not know the particular factors concerned in determining it. Some definite factor or group of factors is of course concerned in each case. Moreover, whenever a gradient originates in response to such action, the specific constitution of the protoplasm at once becomes a factor in determining its further development.

Similar processes of ‘‘adventitious”’ budding are of frequent occurrence in animals, particularly in regulatory development. In short pieces of the naked stem of the hydroid, Corymorpha, the cut ends close rapidly and since there is no perisare there is no such differential exposure as in pieces of Tubularia, but all parts of the surface are exposed alike, unless in contact with the substratum. Such pieces may give rise to single or to biaxial forms like those of Tubularia (See Figs. 93-98) but they often give rise to three or even more axes or partial axes (Figs. 117-121). In such cases, at least one or more of the axes must be adventitious in origin. The buds in these cases represent only the apical regions of individuals but they are of course none the less new axes and must be localized by slight regional differences in cells or cell groups.

Cases like Figures 119-121 show remarkable combinations of extreme apical and extreme basal structures without any intermediate stem region. The basal structure usually appears on that side of the piece in contact with the bottom of the dish, the apical structures at the two cut ends, or if the piece rests on one cut end, this end develops basal structures and the other apical structures. In very short pieces this determination of polarity in relation to the substratum and the differential exposure of the pieces may be quite independent of the original polarity (Child, ’23 c).

Adventitiou3 gradients in embryonic development.—In the embryonic development of the hydrozoa the development of new adventitious axes is readily induced. The normal course of development has already been briefly described (p. 63). When early development occurs under a certain range of inhibiting conditions, e. g., low concentrations of KCN, HCl, or even COs, the blastulze remain spherical, are apparently apolar (p. 82), and in the higher non-lethal concentrations may fail to develop further, even though they remain alive for weeks. With a less extreme degree of inhibition, or if they acclimate to a sufficient degree, they may give rise to new axes, but under

120 12] é Fies. 117-121.—Multiple adventitious polarities in reconstitution of short pieces of Corymorpha stem: Fig. 117, 1/16 piece from tentacles, no stem; Fig. 118, 1/16 piece from 125 mm. stem, biaxial, one axis with two manubria and one set proximal tentacles, other axis with one manubrium; Fig. 119, 1/8 piece from 80 mm. stem, four manubria with distal tentacles and two basal ends; Fig. 120, 1/33 piece from 100 mm. stem, four manubria with distal tentacles, one set of proximal tentacles and one basal end; Fig. 121, 1/23 piece from 120 mm. stem, six manubria with distal tentacles, only one with both distal and proximal tentacles, a

second incomplete set of proximal tentacles below the manubrial complex, one basal region. such conditions the axes are stolons not stems,’ and very commonly not merely one but two or more appear, and in all possible relations to each other (Figs. 122-126). Here again, with the partial or com- * The less specialized hydroid stolon characteristic of this and various other species is a slightly inhibited gradient. Stems can be transformed into stolons by a certain degree of inhibition or depression and transformed back again by removing the inhibiting factor and the embryonic development of stems or stolons can be experimentally controlled in the same way (Child, ’23 b).

_ plete obliteration of the original axis new adventitious axes appear. Under good environmental conditions, 7. e., after removal of the inhibiting agent, all of these new axes may transform into hydranth-bearing stems. Partial or even complete adventitious reduplication of axes is not infrequent in embryonic development, and differences, obviously Figs. 122—126.—Multiple adventitious polarities in inhibited embryonic development of hydrozoan, Phialidium: Figs. 122, 123, development in HCl, m/5000, in the one case four, in the other three stolon axes; Figs. 124-126, development under conditions of crowding and excess of COs, the first with three, the second and third with two stolon axes. In all figures the outer heavier outline indicates perisare, the inner lighter line, surface of coenosarc.

fortuitous, and differing in different embryos, may be concerned in localizing the new axes. Many other cases of partial or complete embryonic reduplication may be called adventitious, which means only that the conditions determining the new gradients and axes are not uniform and constant in all cases, but differ without definite order in different cases, 7. e., are fortuitous. In all cases these adventitious axes are represented by physiological gradients and it is evident that the gradients are not_predetermined but are reactions to local conditions. These duplications,

particularly the partial duplications and partial multiplications such as seen in Corymorpha (Figs. 117-121) are not readily accounted for in terms of inherent molecular or other structural polarity, but they present no difficulties to the conception of polarity as fundamentally a dynamic gradient of organismic magnitude. Dichotomy, twinning and related processes.— Discussion of this group of phenomena is postponed to the following chapter and it need only be noted here that dichotomy and twinning from a single egg consist fundamentally of equal division of a dominant region, or of the replacement of a single dominant region by two of equal rank. The localization of the new axis may be determined by conditions resulting from differential growth in relation to the original gradients, or by conditions in the environment of the dominant region.

The regulatory reduplication of appendages. — The regulatory reduplications of appendages particularly in arthropods and amphibia * afford interesting examples, both of the appearance of new axes — in this case appendage axes or partial axes — and of peculiar axial relations of the reduplicated parts with respect to each other. Considering briefly the case of the amphibian appendages, Harrison’s experiments on transplantation of limb-buds show that in these transplanted buds the antero-posterior pattern is predetermined at the time of transplantation and not altered by the change in position. The dorsoventrality of the leg, however, is not predetermined, but is determined after transplantation by relation to the organismic environment. Expressed in terms of physiological gradients this means that the primary and most strongly marked axial gradient, the antero-posterior gradient, has already determined a persistent physiological differential in the limb-bud, while the less strongly marked dorsoventral gradient has not yet determined a persistent differential. Further experiment may show the possibility of influencing through position on the body both dorsoventral and antero-posterior relations in the limb-bud (see Wilhelmi, 122).

The most interesting features of these transplanted limb-buds are the development of supernumerary limbs by the origin of new limb axes through division or budding of the original axes and the asymmetry relations of the limbs thus formed. One or two supernumerary limbs may arise so that the limb-complex may consist of two or three limbs. Harrison’s statement of the axial relations of such appendages to each other is as follows: (1) The long axes of duplex

or multiplex appendages lie in one plane. (2) Two adjacent members form in structure and position the image of each other, as re- _ flected from a plane mirror bisecting the angle between the respective axes and perpendicular to the common plane of the two axes (Figs. 127, 128). Since the supernumerary limbs are usually of later origin than the original bud it is evident that their asymmetry is determined in some, way in relation to that of the original bud. The asymmetry of the amphibian leg expresses itself in regional differ- _ ences in rate of metabolism and growth and the axial form of the leg is the result of such differences. On the basis of Harrison’s stereochemical hypothesis of symmetry and asymmetry the origin of these regional differences remains obscure; as yet we know nothing of the physiology of this particular process of reduplication except that it is associated in some way with the disturbances brought about by the experiment. Undoubtedly the limb-bud is more or less inhibited for the time being, particularly when grafted into a bodily environment differing from the normal. Under these conditions it must be less effective than the bud in situ in dominating the surrounding region, and physiological isolation with the appearance of new buds may occur, as is probably the case in various other processes of reduplication.

The asymmetry of the supernumerary limb or limbs is obviously determined in some way in relation to that of the primary limb, but until we have more complete knowledge of the processes of reduplication only suggestion is possible. There can be no doubt that differences in electric potential are associated with the regional differences in cellular activity, rate of growth, etc., which constitute the more obvious indications of the asymmetry of the limb. If bioelectric currents of any appreciable strength result from these differences, such currents may ke important or even fundamental factors in determining the relations of the limbs to each other. And in the same way the bioelectric factors associated with the axes of the body on which the limb-bud is grafted may influence its axial relations, at least as regards dorsoventrality.

Recently Przibram (’21) has considered at length a great variety of reduplications of parts in animals and has attempted to interpret their axial relations in terms of a hypothetical space lattice. Such interpretation is purely speculative, absolutely no evidence being presented for the existence of such a space lattice in any case. It seems probable that future experimental investigation will provide a basis for a much simpler physiological interpre-