On Rate Of Movement in a Flat-Worm as Influenced By Body-Level, Duration of Captivity, and Regeneration

Some experience with the Triclad Planarian Polycelis nigra, as the material used in experiments upon behaviour, led the writer to suppose that it would provide a suitable subject for a quantitative enquiry into the relations possibly existing between rate of movement and the intensity of those bodily activities which have been studied by C. M. Child and his school, and have collectively been termed “metabolic rate.”

The investigation here described was begun with the intention of trying to find whether the rapidity of the creeping, gliding motion¹ of these worms differed at various levels of the body in animals cut into three pieces. Planarians are notoriously resistant to the effects of this operation. In these animals local differences in respiratory rate, rate of cell division and growth, and susceptibility to various poisons, are known to bear a fairly definite relation to the longitudinal axis of the body. From the considerable volume of this earlier work, due to Child and his associates, has been evolved an hypothesis of “physiological polarity” which embodies many points helpful in explaining the reactions of at least the simpler organisms. While this hypothesis and its basis are at present not free from criticism, it has appeared to the writer to be worth investigating the problem quantitatively, along the lines indicated above. Additional points of interest in this connection have emerged with reference to the effects of captivity and regeneration upon the rate of movement.

The planarians were collected from one pond ; the first collection was obtained by the writer and taken in a jar with water plants and with an air space above the surface of the water, by road and rail to the laboratory. Subsequent collections from the same pond were despatched by rail under similar conditions. Care was taken that they should not remain longer than was necessary in the closed jars. The dates on which the first, second and third collections were taken were respectively May 17th, June 14th and July 4th, 1928.

The flat-worms were kept in the laboratory in open glass dishes filled with tap-water. The animals were occasionally fed with pieces of earth-worm, which they ate readily ; the duration of captivity on the dates of feeding is given in Tables I and II. As a rule only the main collections were fed, and the remnants of the food were not allowed to remain in the dishes for long; animals under experiment received no food after removal from the stock supply. Feeding was infrequent, but the animals appeared to remain in good condition and were not noticeably more vigorous after food.

Before each experiment a worm was removed from the collection with a widemouthed pipette, placed in a glass dish and the water drawn off ; the arrangement of the eye-spots on the margin of the anterior end was examined with a microscope for purposes of specific identification. The worms were then placed in a separate dish of water. Following identification, the animals were anaesthetised and cut into three pieces. The method used was to place them, by means of a pipette, in a drop of water in a dish filled with paraffin wax ; the water was then withdrawn with a fine pipette and common soda-water substituted for it. This produced anaesthesia and immobility (with some flattening and curvature of the body) which lasted for a short time only. The soda-water was withdrawn and the animal cut transversely with a sharp knife into three pieces of approximately equal length. The pieces were then removed with a pipette, washed in clean tap-water, and heads, midpieces and tails were placed in separate dishes. It is probable that in taking animals from the collection for operation and experiment a size selection occurred, the largest being taken first. This is borne out by examination of the average weights of the successive groups in the tables (I and II), but is of small significance, as will be shown later.

The pieces were left to recuperate from the effects of the operation for a day or so prior to the experiment, the duration of the period being shown in the tables. In some cases all pieces of the same kind were placed in one dish, but in later experiments each piece was placed in a separate dish so that the head, mid-piece and tail of the same animal might be compared as regards their speed.

Five animals were cut at one time, and are considered in the following account as a “group.”

The apparatus used in the following experiments was very simple and consisted of a square dish made from pieces of plate-glass cemented together. The dish was placed over a piece of graph-paper ruled with millimetre squares, which were clearly visible through the plate-glass bottom. An ordinary stop-watch, recording to one-fifth of a second, and a thermometer completed the equipment. The time taken by a piece of worm to travel over a millimetre square was recorded, the observation being made by eye. Attention was concentrated upon the posterior end of the piece.

While the use of a binocular microscope would have been useful for observing movements of the mid-pieces and tails, it would have been impracticable in the case of the heads on account of their speed. Hence, for the sake of uniformity it was not used. Care was taken that the course selected for timing, in each case, was not diagonal to the squares, nor its accuracy vitiated by contraction of the body (as in crawling). Five separate readings were obtained for each piece. The readings for heads, mid-pieces, and tails were recorded separately. The fastest and slowest of each of the three pieces, in any group, did not invariably come from the same worm. The temperature of the water during the trial of each piece was observed. Fresh water was used for each group of five animals.

Unfortunately (but not unexpectedly) the irritability of the heads, mid-pieces, and tails was not uniform, the latter two being far more sluggish than the former, so that it usually became necessary to prod them with a blunt probe in order to induce them to creep.

Each group of pieces was subsequently weighed to about 0·0001 gm., due precautions being taken to avoid errors owing to evaporation or superfluous moisture. The variable period elapsing between testing and weighing may be calculated from the tables. Sometimes the formation and shedding of capsules, or loss of mucus due to occasional accident, may have introduced an error.

The results of these observations of the time taken to cover a distance of 1 mm. are summarised in Tables I and II for Collections 1 (Groups 1–10) and 2 (Groups 11–20) respectively. The temperature range for each group is recorded, together with the duration of their captivity at the time of feeding, operation (cutting), testing and weighing, as well as the average times and average weights for the three pieces separately. The number of trials is given in brackets after the average time¹. In the case of the mid-pieces and tails this time is also expressed as a percentage of the time of the corresponding heads. The mean time over a millimetre for all groups is given for the three pieces, and in the mid-pieces and tails is also expressed as a percentage of that of the heads.

Tables III and IV give the results of combining the time-averages of various groups from Collections r (Groups 1–10) and 2 (Groups 11–20) respectively. The average times, and the duration of captivity when tested, are weighted in the combined groups by the number of trials in each component group of pieces. The percentage proportions of the times of mid-pieces and tails to that of the heads are also given. Further information regarding the combined groups may be obtained from Tables I and II by calculation.

In Table V are given the results of a further test of Collection 1 (Groups 1–10) after they had had time to regenerate their cut tissues to a greater or less extent. The interval elapsing between cutting and this second test is therefore the period of regeneration. Further relevant details are here summarised as in Tables I and II. The animals received no food during regeneration, and in fact were almost entirely neglected, since it was not definitely intended to re-examine them (owing to the impression that continued captivity would make them so sluggish as to be useless). The somewhat unexpected results were therefore obtained after regeneration under the worst of conditions. The only possible source of food was protozoa, sometimes present in the dusty surface film, but that they provided nourishment is very doubtful in view of the fact that the heads and tails, at least, were devoid of a pharynx until a fresh one had regenerated. Moreover, P. nigra is normally a worm of actively carnivorous habit. Nearly all worms had regenerated tissue extending from the line of the original cut, this new tissue being distinguishable on account of its lighter colour. The heads had regenerated tails, the tails new. heads, and the mid-pieces both heads and tails, in the majority of cases. On examination of Group 2, eyes were found in the regenerated heads of the mid-pieces and tails. A few had little or no regenerated tissue, or incomplete extremities¹. Pieces with serious abnormalities or of very sickly appearance were omitted from the results.

Table VI gives an account of various control experiments illustrating certain points of interest. Details are summarised as in the other tables.

The control experiments upon whole worms of Collections 1 and 2 are designed to show the effects of captivity upon normal animals. The worms from Collection 1 were identical in the two trials, though unfortunately they were not tested in the first instance until they had been in captivity some time. The animals of Collection 2 were tested shortly after capture, and again after about a month’s captivity; they were not identical worms.

From Collection 3 large and small worms were selected and tested separately in order to determine the effect, in whole worms, of size upon speed. The worms were selected by eye, and the small animals were, on the average, two-thirds the size of the large ones. They were tested shortly after capture.

Some worms from Collection 3 were tested before and after having their anterior ends split longitudinally for approximately one-eighth to one-fourth the length of their bodies. This experiment was undertaken to determine the effect of injury to the anterior end on rapidity of movement. They were anaesthetised with soda-water and otherwise treated as were the worms in Groups 1–20 (unfortunately the soda-water was stale and in some cases not completely effective). The cuts were not always exactly comparable with respect to length and medial position, but if they were not reasonably similar in nature, the animalswere discarded. An attempt to record speed on the same day as the operation failed on account of continuous writhing. Gliding was comparatively difficult to elicit and usually of short duration.

An experiment upon the effect of increase in intensity of illumination on the speed of the planarians was carried out with animals from Collection 3. Diffuse light fell on the experimental dish from the window in front, while additional light was thrown from two sides (at right angles to the window) by electric bulbs. The artificial light from the two bulbs passed through rectangular troughs containing a solution of alum to cut off the heat generated. The thermometer reading showed this screening to be effective. The bulbs were white and opaque, both 100 volts, one of 100 watts, the other of 150 watts. Their filaments were about 18 inches apart, and the experimental dish lay at an approximately equal distance from either. The speed of each worm was first tested in diffuse daylight alone, and, upon switching on the lamps, the rate under increased illumination was recorded. The intensity of the light from the window and both lamps approximated to that of direct sunlight.

Finally, the possible influence of anaesthesia with soda-water upon rapidity of movement was tested in the case of a few whole worms from Collection 2. The animals were thoroughly anaesthetised (the body becomes much flattened and quite motionless), their speed being tested before, and within half an hour after this treatment. The worms were handled and treated in the same manner as those in Groups 1–20. P. nigra recovers from the effects of soda-water very rapidly, even while still in this medium.

It was found in both collections that the heads were very considerably faster than either of the other pieces, while the mid-pieces were on the whole somewhat faster than the tails. This may be seen in each group from inspection of the average time per millimetre in Tables I and II (also III and IV for combined groups). It will be noticed that in Groups 11–20 (Collection 2) the tails were sometimes faster than the corresponding mid-pieces; in Tables III and IV, which show the average times of the combined groups, this greater rapidity of. the tails than of the mid-pieces is apparent in the last group alone. Fig. 1 shows the mean time for all groups together in Groups 1–10 and 11–20; in both cases the heads are fastest and the tails slowest. It further shows that the mean times of the heads and mid-pieces of Groups 1–10 were less than those of Groups 11–20 (i.e. the former were faster), but that this relation is reversed for the tails. The relative speeds of the two collections are probably intimately related to the duration of captivity of their component groups ; since this was not exactly similar in each case a strict comparison of speed cannot here be stressed. For better comparison of speed in the two collections, the actual average times of the pieces, after varying periods of captivity, may be obtained from Tables I and II. The effect of captivity on change of rate of movement in the two collections is reviewed below. It is sufficient to notice the general and considerable decrease in speed of all pieces, in both collections, with prolongation of captivity.

The proportions that the average times of mid-pieces and tails bear to that of the corresponding head are expressed as a percentage in Tables I and II, for the two collections. This proportion is fluctuating, but its trend is shown in Fig. 2, derived from the data of the combined groups given in Tables III and IV. The curves for pieces from the same collections run more or less parallel, save in the last group of Collection 1 (Groups 1-10) where the percentage of the mid-pieces falls off far more rapidly than that of the tails.

The difference between the curves for Groups 1–10 and 11–20 is referable to the behaviour of their heads, which in Groups 11–20 (see Tables III and IV) shows an increase in average time (a slowing down) far more pronounced than in Groups 1–10; moreover, the mid-pieces and tails of Groups 11–20 do not show such a considerable increase in average time in the final group as do those of Groups 1–10. All this is reflected in a progressively decreasing proportion of mid-piece and tail time to head time in Groups 11–20, while in Groups 1–10 these proportions rise at first and later decrease or remain almost stationary.

It should be remembered that the mean percentage proportions for all groups together are 297 per cent, for mid-pieces and 369 per cent, for tails in Groups 1–10, while in Groups 11–20 the corresponding figures are 217 and 225 per cent. This considerable difference between the percentages of mid-pieces and tails in Groups 1–10, and the much less noticeable difference in Groups 1–20 are well shown in detail by the different combined groups in Fig. 2.

Inspection of the times of both collections in Tables III and IV reveals three main facts. Firstly, that in all three pieces of both collections the average time per millimetre increases directly as the duration of captivity. Secondly, the successively increasing average times of the three pieces do not bear the same relations to each other in Collection 1 and Collection 2 (i.e. the extent of slowing down of the pieces in relation to each other is dissimilar in the two collections). Thirdly, it appears that in the same collection, after the same duration of captivity, the time-increments of the different pieces may be very dissimilar—they have slowed down to a greater or less extent.

The question of rate of slowing down is best discussed in connection with the curves of Fig. 3,¹, which, since they are logarithmic, reduce the successive changes in speed to a common standard. For instance, an increase of time per millimetre from 0·1 to 0·2 sec. would equal an increase from 1 to 2 sec., a doubling of the time in each case. This would be obscured in an ordinary curve by the difference in magnitude of the times concerned. We are here concerned with a change in rate of movement, and, were the rate of the slowing down of any piece constant throughout the experiment, its curve (on a logarithmic scale) would represent a straight line. But it is manifest from Fig. 3 that the rate of slowing down is never constant throughout, and since differences in the slope of the curves indicate differences in rate (the steeper the slope the higher the rate of slowing down), it was found that the rates of the two collections were very different after more or less similar periods of captivity, and, further, that the rates of slowing down of the three pieces of each collection, although synchronising on the whole, vary inter se during the same period (the maximal and minimal rates occur simultaneously, but vary from piece to piece).

The measurement of the angles made by the curves with the abscissa base-line in Fig. 3 provides an index of the rate of slowing down, i.e. the greater the angle the greater is that rate. Table VII gives these indices for both collections, and from it we may see the differences in the behaviour of the pieces as regards slowing down (deceleration), and compare them in the two collections.

In Collection 1 (Groups 1–10) the rate of slowing down is greater in the tails than in the mid-pieces (with the exception of the first part of curve where it is slightly less), and in the mid-pieces than in the heads (an exception being the third part of the curve where the heads’ rate goes up very rapidly). The average rates of heads and mid-pieces are similar, probably owing to the high ultimate rate of the heads ; both are lower than the average rate of the tails.

The rate of slowing down in all three pieces is rapid at first, falls in the second part of the curve, and finally equals, in the mid-pieces, or considerably exceeds its initial rate in the tails and especially the heads.

In Collection 2 (Groups 11–20) the heads always have a rate of slowing down greater than that of the mid-pieces, which in turn have a rate in excess of that of the tails.

The rate of slowing down in all three pieces is high at first (especially the heads), becoming successively lower in the second and third part of the curves, particularly towards the end. It should be particularly noted that the rate of slowing down in the tails has become so low in the last part of the curve, compared to the mid-pieces, that the latter with a higher rate of slowing down have even fallen below the former in actual speed (see Table IV).

In comparing the two collections the first feature that attracts attention is the opposite distribution of high and low rates of slowing down ; in Collection 1 the tails have the highest average rate, and in Collection 2 the heads. Secondly, the average rate of the heads in Collection 2 is higher than in Collection 1, but that of the mid-pieces and tails is lower. The total average for all pieces in Collection 2 is somewhat lower than in Collection 1.

The initial rate of slowing down of the heads is far higher in Collection 2 than in Collection 1, but that of the mid-pieces and tails is lower in Collection 2. In the second part of the curve in Collection 2 a much higher rate is maintained for all pieces than in Collection 1, and this maintenance of comparatively high rate is followed in Collection 2 by a fall to a low rate in the third part of the curve (lower, for all pieces, than in Collection 1), while in Collection 1 the low rate of the second part of the curve gives place to a final high rate of “deceleration.”

These differences in time of maximum and minimum rates of slowing down in two collections would seem to be very suggestive of different inhibitory effects of the environment, whether due to environmental variation or to difference in sensibility. The maximum and minimum rates occur respectively in the third and second parts of the curves in Collection 1, in the first and third parts in Collection 2.

The relation which the rate of slowing down in mid-pieces and tails bears to that of the corresponding heads, in each collection, and at each stage of captivity, may be deduced by examination of curves in which the logarithm of the time of either piece is plotted, in the ordinates, against the logarithm of the time of the corresponding head in the abscissae. Lack of space prohibits the reproduction of these double logarithmic curves, but the data of Table VIII are derived from them. While the percentage curves of Fig. 2 show the proportion which the average times of the pieces bear to that of the heads at each stage (t.e. the reciprocal of the “relative speeds”), and while the rate of slowing down between the corresponding stages may be obtained from Table VII and Fig. 3, the results are best correlated and summarised in Table VIII which gives the “relative decelerations.”

Since the curves from which Table VIII is derived were drawn upon a double logarithmic grid, the angle made between each curve and the abscissa axis is an index of which piece is slowing down most rapidly. If the angle is 45° the rates of deceleration are similar, if it is greater than 45° the pieces represented in the ordinates are slowing down the faster, while if it is less than 45° the heads (in the abscissae) have the higher rate. The rates of deceleration of the two pieces are thus always related to that of the heads (as in the percentage graphs, for the average times) and therefore to each other. If the piece and its corresponding head were to maintain the same relative rates of deceleration throughout, the curve would be a straight line, and the angle would indicate which rate was higher (as above). From Table VIII it is evident that the mid-pieces of Collection 1 have at first a considerably higher rate of slowing down than the heads, in the second part of the curve they almost equal the heads, and finally become considerably lower. The tails, on the other hand, start by being almost as rapid as the mid-pieces, and therefore more rapid than the heads, but continue at a considerably higher rate than the heads in the second part, and almost equal the heads in the final part of the curve. In Collection 2 the rate of slowing down of mid-pieces and tails is always lower than that of the heads. The tails are always relatively lower than the mid-pieces, but both show the same behaviour ; a very low rate in comparison to the heads at first, which increases in the second part of the curve and falls again finally (slightly in the mid-pieces, very markedly in the tails).

The changes in rate of deceleration relative to the heads are thus different in the two collections ; in the first there is a progressive fall from a rate higher than that of the heads to a lower one, in Collection 2 there is a rise from an initial relatively low rate compared to the heads, reaching its maximum in the second part of the curve, and a fall subsequently occurs.

The pieces of Collection 1 have, on the average, a rate of deceleration in excess of, or equal to, that of their heads, while those of Collection 2 have, in every case, a decidedly lower rate than their heads.

An examination of Table V will show that after the pieces of Collection 1 (Groups 1–10) had been allowed to regenerate for periods varying from 23 to 38 days their speed was very considerably increased, the average time for the heads being 0·93 sec. per mm., and that of the mid-pieces and tails 0·63 and 0·69 sec. respectively. Thus after a month, under conditions of regeneration which were far from favourable, the pieces had recovered speeds, in the case of the heads about as high as those shown by the earlier groups shortly after capture (heads in Groups 1-3, Table I), and in the case of the mid-pieces and tails not only higher than the regenerated head pieces, but in excess of the speed shown by any of the worms of the collection. Unfortunately no control tests of the speed of whole worms immediately after capture were made in Collection 1, although this average speed in Collection 2 after 4 days’ captivity was 0·58 sec. per mm. and in Collection 3 after 5 days’ captivity was 0·83–0·84 sec. per mm.¹ These speeds correspond to those of the regenerated mid-pieces and tails. It should be specially noted that whole worms from Collection 1, tested as controls inter 32 days in captivity under more or less the same conditions as the regenerating pieces, showed an average time of 1·16 sec. per mm., or only about half the speed of the regenerated mid-pieces and tails. The single group from Collection 2 (Group 16), which was tested after 18 days’ regeneration, showed much the same characteristics as Groups 1·10, though the tails were here the slowest pieces. The surprising features of this behaviour of the regenerated worms are, firstly, that they should recover so high a speed (in excess of that of controls captive for a shorter time), and secondly, that the regenerated mid-pieces and tails should exceed their heads in speed. It should be noted that the pieces had regenerated their lost tissues in most cases, and that their order of speed appears to coincide with the order of preponderance of new tissue to old, if anterior (head) tissue be considered to play a more important rôle than that of posterior tissue (see discussion below). Thus Table V shows that the mean percentage of the time of mid-pieces to that of the heads is 68 per cent., while that of the tails is 74 per cent, (compare with percentages in Table I) ; the mid-pieces have regenerated both heads and tails, the tails have regenerated anterior tissue (heads) and the heads have regenerated posterior tissue (tails) only. Fig. 1 shows the mean times of the regenerated pieces of Collection 1 graphically contrasted to the pieces before regeneration, and to various controls.

The effect of captivity upon whole worms usually appears to be similar to its effect on worms which are subsequently cut into three pieces; thus Table VI shows that animals from Collection 2 were slower after 34 days in captivity than after 4 days (see Fig. 1); and those of Collection 1 were somewhat slower after 62 days than after 32. This, however, appears to be reversed for Collection 3, where large and small worms were slower after 5 days’ captivity than when tested after 14 and 15 days as “Controls, uninjured” and “Controls, light” respectively— an unexplained result. The slowing down due to captivity is not, however, very pronounced in whole worms.

It may be seen from Table VI that large and small worms, freshly caught, and from the same collection (3), if tested under similar conditions, show similar speeds (see Fig. 1). Size, therefore, does not appear to affect the speed of whole worms. The obvious decrease in size which is visible from the average weights of successive groups in Collections 1 and 2 (Tables I and II) might, however, be correlated with their decrease in speed and both be due to semi-starvation. This, however, is rendered improbable by the behaviour of the whole worms cited above ; further, examination of Tables I and II will show that a weight-decrement series is apparent in successive groups, which, however, were weighed after the same duration of captivity and a similar period without food. The weight-decrement series is probably due to unintentional selection, as each group was made up for testing, of the largest animals from a collection, on account of greater ease of manipulation. This would mask any decrease in weight of successive groups due to semi-starvation, which no doubt occurs to some extent. The time elasping between cutting and weighing does not appear to cause serious loss of weight to the pieces (assuming that the masking effect of selection was invalid), since in the later and lighter groups this period is often less than in the earlier and heavier groups (Table I, and later groups in Table II), and one would expect the weight to be inversely proportionate to the length of this period if injury be the cause of decrement; this, however, is usually not the case. The first six or seven groups in Table II do, however, lend more colour to this view, but it should be noted that in these groups where the period cutting-weighing increases, the period cuttingtesting is very similar, so that loss of weight could not affect the speed in the time available. Finally, pieces allowed to regenerate without food, and therefore presumably decreasing in weight to some extent, are able to recover a high rate of movement. This, together with the similar speed of large and small whole worms, makes it improbable that speed is a function of size.

The effect of temperature variations on the worms’ speed may be gauged from Tables I and II. The temperature range was 6° C. in Collection 1 and over 4° C. in Collection 2. The temperatures rise at first and after a slight fall remain steady in both cases, but finally fall in Collection 1 and rise in Collection 2. This ultimate difference may account for the greater rate of slowing down in the third part of the curve in Collection 1 than in Collection 2. It is curious, however, that the initial period of rapid deceleration should coincide with that of increasing temperature. On the whole temperature variation over the range in question does not seem to affect the speed very markedly and it certainly cannot be said to account for the whole process of loss of speed.

Light-intensity, so far as it was tested, is shown (Table VI) to have small effect upon speed. Thus, worms which showed a speed of 1 mm. in 0·65 sec. in diffuse daylight, when tested in diffuse daylight plus illumination from two powerful electric bulbs, covered the same distance in 0·67 sec.—a closely similar time.

Feeding, which is recorded in Tables I and II, does not appear to affect the speed of the worms permanently, though there appears to be some tendency in both collections for the first group tested after feeding to be faster than that preceding it. This, however, does not delay the general process of loss of speed for long.

The possible effects of anaesthesia upon speed, due to the soda-water used before cutting the worms, is discounted a priori by the extreme rapidity with which they recover from its influence. Table VI shows that the speed of whole worms was slightly greater after treatment than before, probably owing to the unavoidable handling they experienced.

The mechanical stimulation which was necessary to produce movement in some pieces (especially mid-pieces and tails) cannot be quantitatively estimated. It was known to vary from piece to piece as previously mentioned, and it is unfortunate-fortunate that these differences cannot be recorded. It is, however, extremely unlikely that it was responsible for the differences in speed of the three pieces, for the process of slowing down, or for their different rates of deceleration, as this would argue a graduation of stimulation which would be impossibly continuous and exact. It is thus highly improbable that mechanical stimulation with the blunt probe could produce, or seriously alter, the main features of the worms’ behaviour.

In an attempt to discover whether injury to the anterior region, which contains the cerebral ganglia, is equivalent to removing it (as in mid-pieces and tails), the heads of some whole worms were split longitudinally. The speed after the operation was less than half that preceding it. It should be noted, however, that the average time per millimetre of worms with split heads (1·52 sec.) is less than that of mid-pieces and tails in Collections 1 and 2: It therefore appears that injury to the anterior end lowers the speed, but not to so great an extent as its complete removal.

It appears that some of the behaviour described in the preceding section has a resemblance in general outline to the “physiological gradient” phenomena observed by Child and his colleagues in flat-worms and other animals. Several special points deserve mention. In the first place the graduated decrease in speed of pieces, from head to tail, accords well with the known gradients in oxygen consumption and carbon dioxide production, in rates of cell division, growth, differentiation, or reconstitution of lost parts, in which the anterior and posterior ends have respectively the highest and lowest rates (Child, 1924). The metabolic rate of anterior levels of the body appears to be higher than that of posterior ones if we take any of these processes, or speed, as an index of this rate.

Child advances as further evidence for his theory of metabolic gradients in the body (which are supposed to cause, or at least condition, organic polarity, and with this, organic integration) the differences in susceptibility shown by various levels of the body to poisons. The effects of the poison are shown inter alia by actual death and disintegration of the tissues (Child, 1915). These differences in susceptibility are either direct or indirect : in the former case the anterior levels of the body, being more sensitive, die and disintegrate sooner than mid-levels, which in turn do so sooner than the posterior ones. This is the state of affairs in a lethal solution (direct susceptibility) ; in a solution where the concentration of the poison is so low that it can be tolerated to some extent despite certain harmful effects, the order is reversed and the anterior body-levels adapt or acclimatise themselves more readily to the existing conditions than do posterior ones, and survive longer— the hindmost levels doing so least readily (indirect susceptibility). In view of our incomplete knowledge of the conditions governing all activities, including speed, it is unwise to presume further than that the environmental conditions, during and before captivity, may have been sufficiently different in Collections 1 and 2 of polycelis to determine their differences as regards rate of slowing down in the three pieces. If we may consider that the conditions acted to some extent in a manner analogous to the tolerable and intolerable toxic solutions of Child, it would appear that in Collection 2 the environmental conditions during and before captivity were, collectively, more severe than in Collection 1, and were in fact intolerable¹ ; this would react upon the more sensitive anterior levels more strongly than on the posterior ones, and we should not be surprised to find a graduation in the rate of slowing down (an index of unfavourable circumstances) from head to tail. This occurs, the heads slowing down most rapidly and the tails least rapidly. On the other hand, on these premises, the conditions in Collection 1 would appear to be partially tolerable—at any rate less severe than in Collection 2. At first (see Table VII) the conditions caused moderate slowing down, but the heads did so less rapidly than the posterior pieces (i.e. they were more adaptable) ; this state of affairs continued in the mid-part of the experiment (see “second part of curve”), but the conditions were tolerated better by all three pieces and their rate of deceleration fell. Finally, it may be argued, the resistance of the pieces failed under continuance of the unfavourable conditions, and, supporting them no longer, the rate of slowing down increased again in all pieces, but the heads, which are most sensitive once the conditions become intolerable, now had the highest rate. The average rates of slowing down for the three pieces are those one would expect under border-line conditions, the posterior body-level having in this case the highest rate ; the two anterior pieces becoming acclimatised more readily. In Collection 2 the exhaustion, due to severe conditions acting from the beginning of the experiment, might be gauged from the initial rapidity and progressive fall in rate of deceleration from beginning to end in all pieces. In both collections the worms moved extremely slowly in the final stages, especially the mid-pieces and tails. The heads in Collection 1 did not, however, reach so low a speed as those of Collection 2 (Tables I and II) and would no doubt have continued to slow down under conditions that remained unfavourable. Somewhat curiously the tails of Collection 2 never reached so low a speed as those of Collection 1; this should be compared with the rate of deceleration in final stages in Table VII.

If the above interpretation be allowed, the behaviour of the pieces of Collections 1 and 2, as regards slowing down under laboratory conditions, resemble respectively the indirect and direct susceptibility differences as regards poisons, observed by Child.

Regenerated tissue, according to Child (1915), has all the characteristics of young tissue, its metabolic rate resembles the high rate of young, rather than the low rate of old tissue—it appears to be rejuvenated during the process of dedifferentiation, and on subsequent re-differentiation the piece is younger than the worm from which it came. On this basis the recovery of a high speed in regenerated worms appears to depend upon the presence of regenerated tissue in the body. Further, if one may judge from the behaviour of Collection i, the greater the quantity of regenerated tissue the higher the recovered speed, thus the mid-pieces recover the highest speed (see Fig. i), regenerating both heads and tails; while regenerated anterior tissue appears to condition a higher speed than regenerated posterior tissue, for the tails (with regenerated heads) are faster than the head pieces with regenerated tails only. This greater speed of the tail pieces than of the heads is interesting in that it shows that, while both presumably have cerebral ganglia, the regenerated worms with new ganglia (made of regenerated tissue) are faster than those with the original ganglia of the parent worm.

Put shortly, it appears that the recovery of high speed in regenerated pieces of worm is dependent to some extent on the rejuvenation of the body-tissues, and on the relative quantity of young to old tissue, and is not solely dependent on the presence of cerebral ganglia.

It has been shown that control worms (never injured) resemble worms cut before experiment in displaying a decrease in speed with duration of captivity. Perhaps these controls were not starved sufficiently to show the marked increase in activity, which might be expected from the increase in oxygen consumption found in starved worms by Hyman (1919). Starvation appeared to rejuvenate her worms.

The fact that splitting the anterior end decreases speed may indicate that the slowness of mid-pieces and tails relative to heads is due, in part, to anterior injury, but the fact that such posterior pieces are slower than worms with split heads shows that removal of a region with a high metabolic rate has an additionally detrimental effect upon speed. It should be noted that a moderately high speed is maintained despite (presumably) severe injury to the cerebral ganglia. It may be suggested that splitting the anterior end could conceivably lower the rate of vital activities in that region and thus produce slowing down. Any assumptions as to the relative effects of injury to, or total loss of, cerebral ganglia on speed should be received cautiously in view of the behaviour of pieces with regenerated heads; it does not appear that the presence or absence of these organs alone determines the rate of movement. It certainly appears from observation that a split head has its mechanical disadvantages during movement. The speed of regenerated worms is a monument to the hardiness of their species, since they were living under far less favourable conditions, during the process of regeneration, than were the worms in the culture jar from which they were taken ; in point of fact P. nigra tolerates laboratory conditions extremely well, although appearing to creep more and more slowly.

I am indebted to the Council of Birmingham University for monetary support during this investigation. I wish particularly to thank Professor Doris L. Mackinnon for her kindness in allowing me to work in the Department of Zoology at King’s College; Professor J. S. Huxley and Mr C. Roy took a most kindly interest in the work and I am very grateful to them for their helpful advice. Mr D. L. Gunn was good enough to collect and to send to me fresh supplies of material on several occasions.

I have to thank Mr C. F. Hickling for his kindness in reading the manuscript and for very helpful criticism.