1. A method is described by which cells in the zone of outgrowth of cultures in flasks can be photographed at definite intervals over long periods of time.

  2. The rate of growth of chick heart tissue when growing in the presence and absence of embryo extract has been investigated.

  3. A numerical value, the growth index, has been calculated which expresses the mitotic activity of the culture during the periods of observation. In plasma alone the growth index is approximately 3 after 20 hours of growth and falls to zero at the 70th hour if the fluid medium above the plasma (in this case Tyrode) is not renewed. In plasma and embryo extract-(fluid medium embryo extract unchanged) the growth index is approximately 9 after 30 hours of growth, and falls to zero at about the 140th hour. If the fluid medium is frequently renewed with fresh embryo extract the growth index quickly falls to 3, and in the one case followed for more than 4 days it remained fluctuating about that level for the next 6 days, in fact for as long as it was practically possible to prolong the experiment.

  4. It is suggested that the method has many applications in connection with problems associated with the growth rates of tissues and organisms.

In a previous paper (1933) on the growth of tissues in hanging drop cultures, the deficiencies of the method of culture were pointed out, and the desirability of using the flask technique, at any rate for experiments on the influence of the medium on growth, was emphasised. Generally speaking, growth in flasks is very much more uniform than growth in hanging drop cultures, and, since larger volumes of medium can be employed, the environment can be kept relatively much more constant. Consequently from the point of view of obtaining standard tissue upon which to work, flask cultures are greatly to be preferred, especially since tissue which has been acclimatised to life in vitro can be used instead of fresh tissue, and observations can be carried out over very much longer periods. Photographically, however, flasks offer certain pronounced difficulties, only some of which can be overcome. Chief among these is the fact that successful flask cultures demand that the tissues be embedded in a coagulum of plasma. At first sight this seems to offer an insuperable barrier, since the cells grow out into the substance of the clot and consequently occur in several layers and obscure each other in the photographs, but this difficulty can to some extent be overcome by selecting the right consistency of clot, and it is possible by the methods described here to obtain reasonably clear photographs of growth in plasma clots. It should, however, be pointed out that though this overcomes the photographic difficulty it means that growth can only be measured in the presence of plasma, and this, as for example in the case of proteoses, may be undesirable and lead to false conclusions as to the growth-producing capacity of various media. Secondly, growth in flasks must necessarily take place on the bottom of the flask, and the only method of viewing the cells clearly enough for photographic purposes is to observe them from below. This necessitates a reconstruction of the ordinary microscope and the use of flasks with an optical surface at the bottom.

The methods by which successful photographs can be obtained of cultures growing in flasks are described below, together with the results of an investigation of the growth of tissues as it occurs under more or less standard conditions of culture.

The tissue used throughout these experiments has been chick heart tissue taken from 9-or 10-day-old embryos. In some cases it has been used directly and in others the tissue has been subcultured several times in vitro before being used for experiment. The cultures were grown in ordinary Carrel flasks until they were required for photography, when they were transferred to flasks which were specially adapted for this purpose. It was found that the ordinary Carrel flasks could be easily converted in such a way as to make them suitable for use with the camera. By means of small carborundum drills large holes were made in the top and bottom of the flask. The holes were about 2 cm. in diameter, thus leaving but small rims of original glass. These rims were then ground completely flat by the use of carborundum powder on a glass plate, and to them large circular coverslips were sealed by means of shellac. A variety of substances were tried as adhesives, but rather special properties are required. Complete insolubility in water or dilute salt solutions, ability to stand dry heat up to 100 ° C. or more for sterilisation, and stability to dilute salt solutions at 38° C. for prolonged periods, seemed to be the conditions necessary. The most satisfactory results have been obtained by using shellac, after removing as much as possible of its water-soluble constituents which appear to be toxic. When ordinary commercial shellac was used a good joint was obtained which withstood sterilisation at about 90° C. for several hours, but the cultures placed in the flasks failed to show any signs of growth. The commercial shellac was therefore dissolved in alcohol and then precipitated by pouring into a large volume of water. The precipitate was then recollected, washed in water and dissolved in pure 90 per cent, alcohol. It was found that the shellac lost some of its adhesive properties by this method—it lost them almost completely if the process was repeated more than two or three times—but flasks so sealed remained generally intact and cultures grew freely within them. There was no tendency for the fluid medium to creep between the shellac and the glass. Even so it was found advisable to wash new flasks very thoroughly, and latterly they were always incubated with Tyrode solution for some hours before being finally washed in glass-distilled water and sterilised. When all these precautions were taken, flasks with perfect optical surfaces at the top and bottom were obtained, so that critical microscopical observations could be made even under quite high powers.

Previous experiments on the effects of the depth and constitution of the plasma coagulum have shown that in weak and thin coagula the cells spread out over a wider area and become more isolated. This offered therefore the most likely means of obtaining cultures in plasma clots which would be suitable for photographic purposes. After a certain number of preliminary trials it was found that if one part of plasma was diluted with three parts of Tyrode or with two parts of Tyrode and one part of embryo extract (5 c.c. Tyrode per 10-day embryo), a suitable coagulum was obtained. The actual depth of the coagulum varied somewhat owing to the slight variations in the size of the hole which had been cut in the bottom of the flask. The volume used was in nearly every case 0·5 c.c. of plasma, diluted with 0·15 c.c. of Tyrode (or in experiments where embryo extract was being used, with o-1 c.c. of Tyrode and 0·05 c.c. of embryo extract). The clot was therefore as shallow and dilute as it could be if it were still to retain its supporting properties, but in successful cultures the growth was beautifully spread out and the cells clearly separated from one another. Actually these were very nearly all in one plane of focus (16 mm. objective), so that probably they were growing almost exclusively on the glass surface, and only a few cells were finding support in the actual coagulum. The growth remained healthy, and successful subcultures were frequently made from tissue which had been for some time in such media. When the embryo extract was omitted the coagulum was very fragile and one or two experiments were spoilt by the coagulum giving way, and causing the tissue to retract whilst photography was taking place. When the clot had been formed under the influence of the extract this never occurred except after very prolonged growth, but the practice of adding extract to the coagulum is a dangerous one when measurements are to be made of the growth-promoting properties of different media. Some experiments here described illustrate its effects.

The camera used in these experiments was the same as that previously described (1933) but as already indicated the microscope was of a special type constructed for the purpose by Messrs Watson and Sons. It took the form of an ordinary microscope inverted, so that the light and condenser were above the stage, and the objective was below. The light was provided by a 6-volt 24-watt lamp built into the “substage” condenser. The latter was arranged to have a long focus in order to be capable of illuminating the tissue through the whole depth of the flask. For the low-power work described in this paper the condenser is not very important, and satisfactory results have been obtained without its use. The diaphragm was closed down considerably so that as much use as possible is made of the diffraction patterns caused by the edges of the cells, which are otherwise exceedingly transparent The cells also are best kept just slightly out of focus, since they are almost invisible when completely focused, but appear dark when slightly on one side of the critical focus, and then give clear photographic images. The image is obviously not a perfect one, but an imperfect picture for the purposes of counting cells and cell divisions is better than an exact but almost invisible image. The whole “sub-stage” of the microscope can be focused with respect to the stage, and the focusing of the object is brought about by moving the stage and sub-stage together with respect to the objective, which must be kept in a constant position in order to keep the image in the same place on the photographic paper. The latter was again used in preference to film for reasons already mentioned (1933). The camera is not placed directly below the microscope, but the light is deflected at right angles (i.e. horizontally) by means of a prism so that the photographic paper remains vertical. A second prism can be inserted from time to time which will cut off the light from the photographic paper and pass it through an observation eyepiece exactly similar to the one through which it passes on the way to the camera. The two eyepieces are adjustable so that the object can either be sharply focused on the ground-glass screen of the camera or be directly observed without changing the focus. This adjustment is absolutely essential, since the main difficulty encountered in photography of this sort is in the exact adjustment of the focus. As already mentioned the focusing is carried out by movements of the stage relative to the objective, but the critical focus is obtained by a fine adjustment on the objective itself. This was arranged in order to overcome difficulties of a mechanical nature, for the stage and sub-stage together form a rather heavy unit.

The whole microscope was fitted into an electrically heated incubator so that the two eyepieces projected horizontally from adjacent sides. The temperature of the incubator remained constant between 38·5 and 39° C.

Throughout this work the photographs have been taken at 6-min. intervals. This effects a considerable economy in the use of paper and yet does not interfere with the counting of cell divisions. The growth has been estimated as on previous occasions by observing the number of cell divisions occurring in 4-hourly periods and expressing the result as a percentage of the number of cells present on the field during the period, allowance being made for the cells which divide and are therefore not likely to divide again, by subtracting the number of observed mitoses from the final number of cells, i.e. the growth index has been calculated from the expression
where a is the number of cells present at the commencement, b those present at the end, x those that divide during the period, and t the time in hours. As far as possible the value of t has been 4, but owing to technical difficulties in the photographic arrangements periods of greater or shorter duration are sometimes used.

The primary object of these experiments being to obtain a method by which the growth of cells in various media could be measured, the first step was to photograph cultures growing freely in a complete medium and thus to obtain some idea of their behaviour, and to compare the results with those obtained by the use of a medium with limited growth-producing powers. Consequently experiments were made first on tissues growing in plasma and embryo extract, and these were compared with similar experiments using plasma alone. Tissue obtained fresh from embryo hearts has been compared in its behaviour with tissue cultured for more than a month in vitro.

Text-fig. 1 shows the growth index obtained from an experiment using fresh tissue from a 9-day-old embryo. The solid medium, composed of embryo extract and plasma, and the fluid phase of embryo extract remained unchanged from the time the cultures were made till after the end of the period of photographic observation. It is obvious that there are extremely wide variations in the growth index from point to point, but owing to the nature of the material such variations are to be expected since the field of observation contains relatively few cells, generally between 100 and 300, and the number which divide is never very large, so that one or two extra dividing cells during any hour tend to displace that point very considerably. The point marked with a circle is from a similar experiment using a 10-day embryo, and agrees reasonably with those obtained for the 9-day-old chick, but unfortunately the experiment was brought to a premature conclusion, so further points could not be obtained. In these experiments where the medium remained unchanged throughout the photography, it is noticeable that there is a tendency for the growth index to become less as the culture ages. This becomes even more marked in Text-fig. 2 where observations were made over a longer period of time, and the tissue used had been grown in vitro for some time (in one case 2 months) before being used for experiment. With flask cultures where growth is taking place freely, photographs are not satisfactory during the early stages of culture until the cells are reasonably well spread out. When embryo extract was present in the medium photography could rarely be started with profit before the 30th hour. Both fresh and subcultured tissue behave in a similar way. Both show large apparent fluctuations in growth rate. There is perhaps a slight indication that fresh tissue has a somewhat higher growth index, but further experiments are necessary to decide this point. In any case the difference if any is very small and may be neglected for the purposes of the experiments described here, particularly as no differences of any magnitude can be detected when fresh tissue is compared with old tissue in the other media used. The experiments indicate that the growth index attains its highest value soon after the culture is started and falls off steadily, if the medium is unchanged, till it reaches zero, after about 6 days.

Text-fig. 1

Solid phase: 0·5 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. •—• 9-day-old chick heart fibroblasts. Fluid phase: 1 c.c. embryo extract. ⊙ to-day-old chick heart fibroblasts. Fluid phase: 1 c.c. embryo extract renewed once before photography.

Text-fig. 1

Solid phase: 0·5 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. •—• 9-day-old chick heart fibroblasts. Fluid phase: 1 c.c. embryo extract. ⊙ to-day-old chick heart fibroblasts. Fluid phase: 1 c.c. embryo extract renewed once before photography.

Text-fig. 2

Subcultured tissue. Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. ⊙—⊙ Fluid phase: 1 c.c. embryo extract renewed once before photography. •—• Fluid phase : 1 c.c. embryo extract unchanged.

Text-fig. 2

Subcultured tissue. Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. ⊙—⊙ Fluid phase: 1 c.c. embryo extract renewed once before photography. •—• Fluid phase : 1 c.c. embryo extract unchanged.

The next experiment was performed on tissue from a 9-day chick heart, in which the fluid phase of the medium was composed of Tyrode solution instead of embryo extract. The solid phase was composed, as in the previous experiments, of 0·05 c.c. plasma, 0·05 c.c. embryo extract, and 0·1 c.c. of Tyrode. The result obtained (Text-fig. 3) is very similar to those already described. The Tyrode solution was added above the plasma coagulum when this was set, so that there was no washing out of the extract from the clot. This result may be compared with those illustrated in Text-figs. 4 and 5 in which the coagulum was formed without the intervention of extract and the supernatant fluid consisted of Tyrode solution. In these the growth is manifestly far more feeble and ceases after a much shorter time.

Text-fig. 3

9-day-old chick heart fibroblasts. Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Text-fig. 3

9-day-old chick heart fibroblasts. Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract, 0·1 c.c. Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Text-fig. 4

Subcultured tissue. Three experiments. Solid phase: 0·05 c.c. plasma, 0·15 c.c. Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Text-fig. 4

Subcultured tissue. Three experiments. Solid phase: 0·05 c.c. plasma, 0·15 c.c. Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Text-fig. 5

9-day-old chick heart fibroblasts. Solid phase: 0·05 c.c. plasma, 0·15 Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Text-fig. 5

9-day-old chick heart fibroblasts. Solid phase: 0·05 c.c. plasma, 0·15 Tyrode solution. Fluid phase: 1 c.c. Tyrode solution unchanged.

Again, Text-fig. 4 can be compared with Text-fig. 5, and very little difference in behaviour is noticeable between fresh tissue (9-day old) and tissue which had been repeatedly subcultured.

Text-fig. 6 is a composite figure obtained by grouping together all the values of the growth index obtained when embryo extract formed part of the medium, and all those for cultures grown in the absence of extract. The medium has not been changed in any of the cultures, and although there is a considerable range of variation the two sets of points fall absolutely clearly into two groups. Embryo extract increases the growth index and a culture continues to grow for approximately twice as long in a medium containing extract, and from Text-fig. 3 it is obvious that a comparatively small concentration of extract in the coagulum is sufficient to produce a great increase in the growth capacity of a culture.

Text-fig. 6

medium containing no embryo extract; ⊙ medium containing embryo extract.

Text-fig. 6

medium containing no embryo extract; ⊙ medium containing embryo extract.

This very clear separation of the points representing the growth rate in different media promises well for the use of the method here described in studying the effects of so-called growth-promoting substances.

In the normal method of growing tissues in flask cultures the fluid medium is renewed every 2 days. It was therefore thought to be interesting to obtain the mitotic index on each day with this type of culture, and to compare it with those found in the previous ones where the fluid had remained unchanged. Since the cultures may last up to 10 days or more, it became impracticable to photograph them continuously, so they were photographed for an 8-hour period each day. Actually this proved rather unsatisfactory, as it was difficult to distinguish between possible rhythms and normal variation, because as the culture aged the variation actual or artificial reached large dimensions. Also a suitable technique has not yet been established for changing the medium satisfactorily without disturbing the culture flask and thereby causing possible fluctuations. By far the greatest difficulty, however, lay in the annoying property possessed by embryo extract of depositing a fine precipitate, presumably of denatured protein, on standing. Some cultures became impossible to photograph after two or three changes owing to this precipitate, which completely obscured the images of the cells. An attempt was made to overcome this difficulty by allowing the extract to stand for some hours before use and centrifuging it immediately before adding to the culture. It is at present uncertain how far this is justifiable, since it is possible that some of the growth-producing substances may be carried down and lost with the precipitate. From the figures obtained there is no positive evidence that this is the case, since cultures treated with recentrifuged extract show as high values as those with normal extract, but comparatively few figures have so far been available.

The growth indices for cultures in which the fluid has been repeatedly renewed are given in Text-fig. 7. Owing to the difficulties outlined above only one culture has been successfully photographed for any great length of time, and that shows great fluctuations. However, from these few rather unsatisfactory data two points would seem to stand out quite clearly. First, cell division is maintained in the culture for a very much longer time when the medium is renewed, and secondly, the growth index has a lower value. In an unchanged medium the growth index falls fairly steadily from say 10 (by extrapolation) through about 5 after 60 hours to zero after about 6 days. If the medium is renewed the index falls almost at once to a value in the neighbourhood of 3 and apparently stays there for a considerable time, e.g. in the one case followed, for nearly 10 days. This is interesting in that it suggests that the plasma plays some important part in stimulating the intense growth which occurs in the early stages of an unchanged medium, and that by washing the plasma this important factor is removed, so that after the first one or two renewals of the fluid medium the culture behaves as if growing in embryo extract alone, except for the mechanical support of the clot. In hanging drop cultures in embryo extract alone after 30 hours the growth index was about 4·5 and dropping quickly, which corresponds reasonably with the points for that period shown in Text-fig. 7, due allowance being made for the different conditions of culture. It is noticeable that the growth index reaches its highest values in the one experiment followed some 15-20 hours after the fluid medium was renewed. The data are obviously insufficient to attach much weight to them, but it seems possible that by this prolonged method of cultivation the growth actually takes place in a series of waves occurring fairly regularly after each change of medium. This does not mean that the cells of the culture show rhythmic divisions, but that the conditions for growth are improved each time the medium is changed and deteriorate again before the next renewal.

Text-fig. 7

Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract and 0·1 c.c. Tyrode solution. Fluid phase: 1 c.c. embryo extract, frequently renewed. Curve showing approximate values of growth index in unchanged extract. - - - - Approximate values of growth index in the absence of embryo extract. ⊙ Points from one experiment in which the extract was renewed at the points marked by arrows.

Text-fig. 7

Solid phase: 0·05 c.c. plasma, 0·05 c.c. embryo extract and 0·1 c.c. Tyrode solution. Fluid phase: 1 c.c. embryo extract, frequently renewed. Curve showing approximate values of growth index in unchanged extract. - - - - Approximate values of growth index in the absence of embryo extract. ⊙ Points from one experiment in which the extract was renewed at the points marked by arrows.

Since it is not possible, owing to the extension of the culture, to keep the same field under the camera continuously by these methods, little numerical information of value can be given as to the extent to which the total increase in the number of cells is taking place. In the experiment which lasted for nearly 10 days cells were still wandering on to the field at the end but at a slower rate than earlier, and at that time practically the whole of the “growth” could be accounted for by cell division. This probably explains the observed phenomenon that flask cultures apparently stop growing after some days. When this occurs the culture is probably spread out as far as is necessary to eliminate any steep gradient of metabolites outwards from the central implant, so that the cells move slower and less centrifugally. Any further increase in the size of the colony is explicable simply on the ground that room must be made for newly produced cells whose movements are becoming more random. In the periphery the cells double their number in about 33 hours, and in the light of the experiments previously described the rate of division in the periphery of a culture is greatly in excess of the figure for the rate of increase where the cells are more crowded and still more in excess of the figure for the central implant (Fischer and Parker, 1929), and consequently the rate of increase of the culture as a whole is probably very much slower, after active outward migration has ceased, being dependent almost entirely upon the multiplication of the cells in the periphery. Cultures made in a medium composed as described above, however, although they grow well and in a healthy manner for some time, do not survive for very long, since the enzymes in the extract slowly liquefy the plasma, and this alone makes it necessary to subculture them fairly frequently and more often than would be necessary to keep up a constant rate of growth. It would be interesting to obtain records of the growth of tissues kept for a longer time in a more solid medium, but here the difficulties of photography are increased.

Besides the quantitative differences in growth rate just described the general character of the cells varies somewhat according to the medium. In cultures grown in the entire absence of embryo extract the cells remain relatively radially arranged with respect to the central implant. They are always close together and typically spindle-shaped. A few on the extreme periphery show an inclination to separate and wander into the medium. The peripheral cells are nearly always larger than those more centrally placed. This is true of all media observed. The crowding together of the cells and their spindle shape are to some extent accounted for by the slow rate of migration. The average rate of migration of several cells followed during the second day of growth in cultures without extract worked out at 11·2μ per hour, whereas in cultures of a similar age but made with embryo extract the average rate of migration was 42·7 μ per hour. If this figure is compared with the similar one (12 μ) calculated for the cells growing in a hanging drop of embryo extract and reported in a previous paper (1933) it is at once clear how important a part is played by the plasma in assisting migration, and this is probably not all to be explained on the grounds of its mechanical properties, since the rate of migration in plasma alone is slow. It is slow too in extract alone, but in the combination of the two it becomes at once very much higher. In hanging drop cultures too a high rate of migration was observed in cultures in a mixture of serum and extract, so that at least some of the beneficial effects of the plasma lie in its liquid part. The behaviour of the cells suggests that plasma provides mechanical support and contains food substances which are, however, not readily accessible to the cells. Serum contains these same food substances. Embryo extract contains certain, probably other, nutritive bodies which are immediately available to the cells and also contains the means of liberating the food substances of plasma and serum and making them accessible.

The cells which surround a culture growing in plasma alone are spindle-shaped and extremely hyaline. But in cultures containing embryo extract they show a much more varied appearance. They scatter more freely in the medium, lose their definite orientation and show considerably more fat globules. Although the cells on the periphery are again large as compared with those near the centre, they do not give the appearance of containing so much cytoplasm as cells similarly situated in plasma cultures, particularly as part of their substance is taken up by the fat globules.

The photographs on Pl. I are all of the same culture, during its 4th day of growth. Between Fig. 1 and Fig. 2 there is a lapse of 6 hours, and the culture medium is composed of a solid phase of plasma and Tyrode solution and a fluid phase of Tyrode solution. Between Figs. 2 and 3 the Tyrode solution of the fluid phase was replaced by embryo extract. Fig. 3 shows a slightly different field of cells, but Fig. 4 shows the same field after the lapse of another 6 hours. The difference in the appearance and the rate of movement in the two media is very conspicuous.

The photographs on Pls. II and III show the periphery of a typical culture in plasma alone, and that of a similar culture fed with extract, each taken at 6-min. intervals.

The duration of mitosis remained constant in cultures growing in embryo extract until growth had almost ceased, when dividing cells often became fixed in metaphase or late prophase for some time, before producing two daughter cells. In plasma cultures the whole process of mitosis was often somewhat prolonged, so that instead of being over in about 30 min. it lasted for something over 36 min. and in several cases for as long as 60 min. Here it was not a question of the cell becoming fixed in one phase of division, but the whole process appeared to go slower, just as cells in plasma move more slowly than those with embryo extract.

The cultivation of cells in vitro offers an ideal method for the study of the process of growth, and since the conditions can to some extent be experimentally controlled quantitative data should be obtainable. Hitherto methods of measuring growth have been almost confined to estimating the increase in area of cultures, and to estimating the division rate of cells at fixed times after the making of the culture, a process in which the necessary staining ends the life of the culture. In the almost complete absence of measurements of the increase in weight of the cultures which are technically very difficult to obtain, the most satisfactory method of measuring growth is by estimating the proportion of cells which are dividing, in other words determining what may be called the growth index of the tissue. This growth index has been estimated on fixed preparations by various authors, and recently Olivo and Delorenzi (1932) have published the results of direct continuous observations on the rate of growth of the periphery of tissue cultures in plasma and embryo extract, and have worked out the growth index in each case. The basis upon which this has been calculated is not quite the same as the one adopted here. Olivo and Delorenzi calculated the proportion of cells per thousand which are observed to be in a state of division at any one time. The basis for calculation for the present investigation has been the percentage of cells which show telophase (i.e. actual division into two) per hour. The latter is adopted in order to overcome any errors introduced by variations in the actual time of division. In embryo extract and plasma a cell takes approximately half an hour to divide, so that if the growth indices obtained by Olivo and Delorenzi are divided by 5, they become almost exactly comparable to those found in the present investigations. This, however, only applies to cultures in plasma and extract where the time of division remains constant at half an hour. From these figures it is possible to calculate the interkinetic period, i.e. the time between successive divisions of a cell. This calculated value however does not give any further information than the growth index, and it is doubtful whether measurements of the interkinetic periods of the cells of a culture mean very much. Olivo and Delorenzi from direct observations showed that this might be 7 hours or 21 hours in cells in the same field, and similar wide variations have been observed in the present series of experiments. A tissue as a whole may show an average interkinetic period which may be calculated from its growth index, but this tells little as to how often individual cells are dividing. In fact it seems probable that, in any given tissue, some cells divide fairly frequently while others only divide comparatively rarely. Whether this would become more constant the longer the tissue was cultured in vitro is a point worthy of investigation, and would show whether a uniform strain of tissue was really being obtained. The experiments described here indicate that prolonged cultivation does not lead particularly to uniformity in the value of the growth index. The work of Olivo and Delorenzi indicates that widely different interkinetic periods occur in tissue which has been subcultured more than 2000 times, and in the tissues of the body there must exist extremely different rates of growth even in different parts of the same tissue, all of which cannot be explained as being produced by differences in the environment : in other words they are probably inherent in the cell ; some cells have a high growth rate, others a low one, and these differences being part of the essential make-up of the cell probably persist even into the foreign environment of a tissue culture. If the conditions of the culture are favourable to growth then the cells will grow and divide each at their own rate, and if the conditions are less growth-promoting, the cells remain more or less quiescent, but any growth which occurs would show the same relative rates between the individual cells.

The methods of investigation described here now give a means of obtaining some direct comparison between the behaviour of cells in various media. Their growth rate can be observed continuously, and any changes in the activity and morphology of the cells become at once evident. Moreover, the calculation of the growth rate is relatively independent of the disturbing influence of the spreading out of the initial implant of the culture. Although considerable fluctuations occur in the growth index, which might vitiate results calculated from observations made at any one time, the fact that the observations can by the present method be continued for long periods on the same culture makes the method at once more reliable. It is true that there is known to be a considerable difference in the behaviour of cells in plasma according to whether embryo extract is present or not, but it is very striking that in the figure showing the growth indices at different times for cultures in these media there is a complete separation of the points for the two media. The method therefore should prove useful in attempting to answer the vexed question as to whether or not it is possible to stimulate cells to divide by hormone-like bodies.

By performing a series of experiments on tissue of different ages it should be possible to investigate whether in the course of embryonic development there are variations in the growth rate of particular organs. The work of Schmallhausen (1926) and others suggests for example that in the chick the growth rate of certain organs is high on the 9th and 10th days, low on the 11th day, and that there is again a peak in the growth-rate curve on the 12th and 13th days. It would be interesting to see if these outbursts of growth were caused by extra cell divisions, and if so if they would be carried over by tissues growing in vitro. Actually it has been observed in previous experiments that 12-day-old chick heart tissue shows an abundance of mitoses and seems to grow by this method more freely than 10-day-old tissue, although it is less active in spreading out from the implant. The data are, however, only suggestive and much further work will have to be done to determine whether such is actually the case. If so the method should prove useful in determining the actual growth rate of tissues at various times in the development of the organism, and possibly help to clear up some of the problems of the relative growth of parts of an organism, such as whether the rates of growth are intrinsic in the tissues, or whether they are impressed on them by bodies circulating in the organism, and if the latter, at what time the growth rate becomes affected.

When chick heart tissue is cultured in embryo extract, and the medium is frequently renewed, the growth index after a time fluctuates about a mean of 3 per cent. ; it would be interesting to investigate whether other tissues grown under similar conditions grow at a uniform rate and whether this also would give a growth index of 3 per cent. Is 3 per cent, more characteristic of the medium or of the tissue?

During the course of the experiments, as will occur in tissue culture, some of the cultures became infected. One became infected on the 4th day of cultivation after the fluid medium (embryo extract) had been renewed for the second time, and it so happened that this culture had been photographed each day. On the 4th day it was observed to be infected and was unfortunately discarded at once, but the photographs were developed and showed that during the last 8 hours cell divisions had been occurring at an inordinately high rate, giving a growth index of 12 per cent. The nature of the infection was not identified. It took the form of a cloudiness confined to the area of the coagulum immediately in contact with the tissues. A similar “growth” has been observed since in other cultures, but has so far not responded to the efforts of bacteriologists to make cultures of it, and so is still unidentified. There seems to be no other explanation of the greatly increased growth rate except that it was in some way connected with the appearance of the infection.

The method here described, therefore, should be useful in an analysis of a variety of problems connected with the rate of growth of tissues, and might profitably be applied not only to normal tissues, but also to neoplasms.

The expenses of this research have been in part defrayed by a grant from the British Empire Cancer Campaign. The author is greatly indebted to the Director ofthe Strangeways Research Laboratory for facilities for work during building operations in the Physiological Laboratory, and for the supply of plasma for the experiments. Grateful thanks are also due to Dr F. G. Spear for helpful criticism both during the work and in the preparation of the manuscript.

Fischer
,
A.
and
Parker
,
R. C.
(
1929
).
Brit. Journ. Exp. Path
.
10
,
312
.
Olivo
,
O. M.
and
Delorenzi
,
E.
(
1932
).
Arch.f. Exp. Zellforschung
,
13
,
221
.
Schmallhausen
,
I.
(
1926
).
Roux Arch.f. Entwicklungs. d. Organismen
,
108
,
322
.
Willmer
,
E. N.
(
1933
).
Journ. Exp. Biol
.
10
,
317
.

Plate I

For explanation see text, p. 350.

For explanation see text, p. 350.

Plate II

Six consecutive photographs of a culture growing in plasma alone, with Tyrode as the fluid medium. One cell is undergoing mitosis.

Six consecutive photographs of a culture growing in plasma alone, with Tyrode as the fluid medium. One cell is undergoing mitosis.

Plate III

Six consecutive photographs of a culture growing in plasma and embryo extract, with embryo extract as the fluid medium. Note the frequent mitoses.

Six consecutive photographs of a culture growing in plasma and embryo extract, with embryo extract as the fluid medium. Note the frequent mitoses.