in longitudinal, not in general extensibility. I presume that these writers might say that the excess in longitudinal extensibility is always present whether general extensibility is greater or less. In the meanwhile we must pass on to more recent researches connected with surface-growth by apposition.

In Strasburger's later work, 'Histologische Beiträge,' 1889, his views on growth have undergone considerable modification. The study of certain epidermic cells, of the folds in membranes, and the repetition of Krabbe's work on certain bast fibres, have convinced him that apposition does not account for all forms of growth. Krabbe showed that in full-grown sclerenchyma (e.g. in Oleander) local widenings occur without any such amount of thinning in the membrane as would occur if the bulging were due to stretching. The only possible explanation seems to be that there is a migration of new material into the cell-wall. Such intussusception might be, as Nägeli supposed, a flow of fluid out of which new micellæ crystallise; but it is now established that cellulose arises as a modification of protoplasm, so that it would harmonise with our knowledge of the origin of cellulose if we assume that intussusception was preceded by a wandering of protoplasm into the cell-wall. Such a state of things would render possible the regulation of longitudinal growth in the case of Nitella and Spirogyra, already alluded to, as well as in growth curvatures. This view might also harmonise with Wiesner's theory that the cell-wall contains protoplasm as long as it continues to grow.

For the sake of brevity I content myself with the above examples: I think it will be allowed that there is a focussing of speculation from many sides in favour of 'active' surface-growth-or, what is perhaps a better way of putting it, in favour of a belief that the extension of cell membranes depends on physiological rather than physical properties, that it is in some way under the immediate control of the protoplasm. We may take our choice between Wiesner's wallprotoplasm (dermatoplasm), protoplasmic intussusception as conceived by Strasburger, or the action of the ectoplasm in the manner suggested by Vines,3 who supposes that the crucial point is a change in the motility of the protoplasm, not of the cell membrane. The latter theory would undoubtedly meet the difficultiesif we could believe that so yielding a substance as protoplasm could resist the force of turgor.

The great difficulty is, it seems to me, that since e.g. in Caulerpa, surface-growth is clearly due to stretching, as Noll has demonstrated, and since in osmotic cell-pressure a stretching force does exist, it cannot be doubted that turgor, and ordinary physical extensibility are conditions of the problem. This remains true in spite of Klebs' curious observations on the growth of plasmolysed algæ, or in spite of the fact that pollen tubes may grow without turgor, in spite of the same being perhaps true of young cells filled with protoplasm. In the face of all these facts, osmotic pressure in the cell must remain a vera causa tending to surfacegrowth.

If we accept some form of active' surface-growth, we must deal with turgor in another way, although to do so may require a violent exercise of the imagination. Are we to believe, for instance, that the function of turgescence is the attaining of mechanical strength? If we hold that cell-walls increase in area independently of turgor, we shall be forced to invent a hypothesis such as the following which I am far from intending to uphold. It is possible to imagine that the function of the force of turgor is merely to spread out the growing membrane to its full extent, and, as it were, to make the most of it. Turgor would in this respect play the part occupied by the frame used in embroidery, making it easier to carry on the work satisfactorily, but not being absolutely necessary. When

1 Pringsheim's Jahrb. xviii.

2 Sitz. Wien. Akad. 1886, vol. xciii. p. 17.

Sachs' Arbeiten, 1878, and Physiology, 1886. See also Gardiner, on protoplasmic contractility, in the Annals of Botany, i. p. 366. Pfeffer has, I think, shown that Vines' and Gardiner's theories assume the existence of too great strength in the ectoplasm. See Pfeffer in Abhandl. der k. Sächs. Gesellsch. xvi. 1890, p. 329.

Tübingen. Untersuchungen, ii. p. 489.

See Noll, Würzburg. Arbeiten, iii. p. 530.

mechanical strength is gained by turgor (as in Mucor), instead of by brute strength of material, as in a tree-trunk, a great economy in cellulose is effected. If turgor played our hypothetical part of smoothing out the membrane and insuring that it shall occupy as large a space as possible, it would effect the same kind of

economy.

It is not necessary to inquire how far this hypothesis accords with our knowledge of cell mechanics. It is only put forth as an example of the difficulties in which we land if we seek for a new function for turgor. We are, indeed, surrounded by difficulties; for, though the theories which are classed together as protoplasmic have much in their favour, they, too, lead us into an impasse.

Circumnutation.

I shall conclude by saying a few words about the theory of growth-curvatures put forward in the Power of Movement in Plants.' I can here do no more than discuss the relation of circumnutation to curvature, which is the thesis of the book in question, without attempting to enter the arena with regard to the many objections which have been raised to other parts of our work.

A distinguished botanist, Professor Wiesner, of Vienna, published in 1881 a book, 'Das Bewegungsvermögen der Pflanzen,' entirely devoted to a criticism of the Power of Movement.' It is founded on a long series of experiments, and is written throughout in a spirit of fairness and candour which gives it value, apart from its scientific excellence, as a model of scientific criticism. The words written on the title-page of the copy presented to my father are characteristic of the tone of the book:-'In getreuer Opposition, aber in unwandelbarer Verehrung. A letter printed among my father's correspondence shows how warmly he appreciated his opponent's attack both as to matter and manner. Wiesner's opposition is far-reaching, and includes the chief theoretical conclusion of the book, namely, that movements such as heliotropism and geotropism are modifications of circumnutation. Neither will he allow that this revolving nutation is the widely-spread phenomenon we held it to be. According to Wiesner, many parts of plants which do not circumnutate are capable of curving geotropically, &c.; he is, therefore, perfectly justified, from his own point of view, in refusing to believe that such curvatures are derivations from circumnutation. He points out that our method of observing circumnutation is inaccurate, inasmuch as the movement is recorded in oblique projection. This we were aware of,' and I cannot but think that Wiesner has unintentionally exaggerated its inaccuracy; and that, if used with reasonable discretion, it cannot lead to anything like such faulty records as in the supposititious cases given by our critic. However this may be, Wiesner's results are perhaps more trustworthy than ours, and should receive the most careful consideration.

Wiesner's conclusions, taken from his own summaries, are as follows:

The movement described as circumnutation is not a wide-spread phenomenon in plants. Stems, leaves, and acellular fungi are to be found which grow in a perfectly straight line. Some roots grow for considerable periods of time without deviating from the vertical. When circumnutation does occur, it cannot be considered to have the significance given to it in the Power of Movement.' The movements observed by Wiesner are explained by him in three different ways:

i. As the expression of a certain irregularity in growth depending on the want of absolute symmetry in structure, and on the fact that the component cells of the organ have not absolutely similar powers of growth.

ii. As the expression of opposing growth-tendencies. Thus certain organs have inherent tendencies to curve in definite planes-for instance, the bending of the hypocotyl in the plane of the cotyledons. Wiesner believes that such tendencies, when combined with others-heliotropic, geotropic, &c.-lead to alternate bendings in opposite directions, according as one or other of the components is temporarily the stronger.

iii. Wiesner allows that circumnutation does exist in some cases. This last

1 Power of Movement, p. 8.

class he considers a small one; he states, indeed, that nearly all, especially the clearly perceptible circumnutations,' are combined movements belonging to the second of the above categories.

Although I have perhaps no right to such an opinion without repeating Wiesner's work, yet I must confess that I cannot give up the belief that circumnutation is a widely-spread phenomenon, even though it may not be so general as we supposed.

If, then, circumnutation is of any importance we are forced to ask what is its relation to growth-curvatures. It was considered by my father to be the basis or groundwork for the acquirement, according to the requirements of the plant, of the most diversified movements.' He also wrote: 2 A considerable difficulty in the way of evolution is in part removed, for it might be asked how did all these diversified movements. . . . first arise? As the case stands, we know that there is always movement in progress, and its amplitude, direction, or both, have only to be modified for the good of the plant in relation to internal or external stimuli.'

Those who have no belief in the importance of circumnutation, and who hold that movements may have arisen without any such basis, may doubtless be justified in their position. I quite agree that movement might be developed without circumnutation having anything to do with the matter. But in seeking the origin of growth-curvatures it is surely rational to look for a widelyspread movement existing in varying degrees. This, as I believe, we have in circumnutation: and here comes in what seems to me to be characteristic of the evolution of a quality such as movement. In the evolution of structure, each individual represents merely a single one of the units on which selection acts. But an individual which executes a number of movements (which may be purposeless) supplies in itself the material out of which various adapted movements may arise. I do not wish to imply that tentative movements are of the same order of importance as variations, but they are undoubtedly of importance as indication of variability.

The problem may be taken back a stage further; we may ask why circumnutation should exist. In the 'Power of Movement' (p. 546) we wrote: Why every part of a plant whilst it is growing, and in some cases after growth has ceased, should have its cells rendered more turgescent and its cell-walls more extensile first on one side then on another... is not known. It would appear as if the changes in the cells required periods of rest.' Such periods of comparative rest are fairly harmonious with any theory of growth; it is quite conceivable by intussusceptionists and appositionists alike that the two stages of elongation and fixation should go on alternately, but this would not necessarily lead to circumnutation. It might simply result in a confused struggle of cells, in some of which extension, in others elongation, was in the ascendant; but such a plan would be an awkward arrangement, since each cell would hinder or be hindered by its neighbour. Perfection of growth could only be attained when groups of contiguous cells agreed to work together in gangs, that is, to pass through similar stages of growth synchronously. Then, if the different gangs were in harmony, each cell would have fair play, elongation would proceed equally all round, and the result would be circumnutation.4 Whether or no any such origin of circumnutation as is here sketched may be conceived, there can be no doubt that it had its origin in the laws of growth apart from its possible utilisation as a basis for growth-curvature.

It is, however, possible to look at it from a somewhat different point of view, namely, in connection with what Vöchting has called rectipetality. He made out the fact that when an organ has been allowed to curve geotropically, heliotropically, &c., and is then removed from further stimulation by being placed on

1 Power of Movement, p. 3.

2 Loc. cit., p. 4.

* Strasburger, Histolog. Beiträge, p. 195, speaks of the pause that must occur after the formation of a cellulose lamella. Hofmeister, Württemburg. Jahreshefte, 1874, describes the growth in length of Spirogyra as made up of short intervals of rapid growth alternating with long pauses of slow growth.

• I purposely omit the circumnutation of pulvini. • Die Bewegung der Blüthen und Früchte, 1882.

the klinostat, it becomes straight again. This fact suggested to Vöchting his conception of rectipetality, a regulating power leading to growth in a straight line. It may be objected that such a power is nothing more than the heredity, which moulds the embryo into the likeness of its parent, and by a similar power insists that the shoot or root shall take on the straight form necessary to its specific character. But the two cases are not identical. The essence of rectipetality is the power of recovering from disturbance caused by external circumstances. When an organ has been growing more quickly on one side than another, the regulating power reverses this state of things and brings the curving organ back towards the starting-point. We have no means of knowing how this regulating power acts in undisturbed growth. It is possible to imagine a type of irritability which would insure growth being absolutely straight, but it is far more easy to conceive growth as normally made up of slight departures from a straight line, constantly corrected. In drawing a line with a pencil, or in walking towards a given point, we execute an approximately straight line by a series of corrections. If we may judge in such a matter by our own experience, it is far more conceivable that the plant should perceive the fact that it is not growing absolutely straight and correct itself, than that it should have a mysterious power of growing as if its free end were guided by an external force along a straight-edge. The essence of the matter is this: we know from experiments that a power exists of correcting excessive unilateral growth artificially produced; is it not probable that normal growth is similarly kept in an approximately straight line by a series of aberrations and corrections? If this is so, circumnutation and rectipetality would be different aspects of the same thing.

This would have one interesting corollary: if we fix our attention on the regulating power instead of on the visible departures from the straight line, it is clear that we can imagine an irritability to internal growth-changes existing in varying intensities. With great irritability very small departures from the straight line would be corrected. With a lower irritability the aberration would be greater before they are corrected. In one case the visible movement of circumnutation would be very small, in the other case large, but the two processes would be the same. The small irregular lateral curvatures which Wiesner allows to exist would therefore be practically of the same value as regular circumnutation, which he considers comparatively rare.

The relation between rectipetality and circumnutation may be exemplified by an illustration which I have sometimes made use of in lecturing on this point. A skilful bicycle-rider runs very straight, the deviations from the desired course are comparatively small; whereas a beginner wobbles' or deviates much. But the deviations are of the same nature; both are symptoms of the regulating power of the rider.

We may carry the analogy one step further: just as growth-curvature is the continuance or exaggeration of a nutation in a definite direction, so when the rider curves in his course he does so by wilful exaggeration of a 'wobble.'

It may be said that circumnutation is here reduced to the rank of an accidental deviation from a right line. But this does not seem necessarily the case. A bicycle cannot be ridden at all unless it can 'wobble,' as every rider knows who has allowed his wheel to run into a frozen rut. In the same way it is possible that some degree of circumnutation is correlated with growth in the manner suggested above, owing to the need of regular pauses in growth. Rectipetality would thus be a power by which irregularities, inherent in growth, are reduced to order and made subservient to rectilinear growth. Circumnutation would be the outward and visible sign of the process.

I feel that some apology is due from me to my hearers for the introduction of so much speculative matter. It may, however, have one good result, for it shows how difficult is the problem of growth-curvature, and how much room there still is for work in this field of research.

The following Reports and Papers were read :

1, Fourth Report of the Committee appointed for the purpose of reporting on the present state of our knowledge of the Zoology and Botany of the West India Islands, and taking steps to investigate ascertained deficiencies in the Fauna and Flora.-See Reports, p. 354.

2. Report of the Committee appointed to report on the present state of our knowledge of the Zoology of the Sandwich Islands, and to take steps to investigate ascertained deficiencies in the Fauna.-See Reports, p. 357.

3. Fifth Report of the Committee appointed for the purpose of taking steps for the establishment of a Botanical Laboratory at Peradeniya, Ceylon.See Reports, p. 358.

4. Report of a Committee appointed to make a digest of the observations on the Migration of Birds at Lighthouses and Light-vessels which have been carried on by the Migration Committee of the British Association.-See Reports, p. 363.

5. Fourth Report of the Committee for the purpose of collecting information as to the Disappearance of Native Plants from their Local Habitats.— See Reports, p. 359.

6. Report of the Committee appointed for the purpose of arranging for the occupation of a Table at the Laboratory of the Marine Biological Association at Plymouth.-See Reports, p. 364.

7. Report of the Committee appointed for Improving and Experimenting with a Deep-sea Tow-net.-See Reports, p. 382.

8. Non-sexual Formation of Spores in the Desmidiaceœ.
By A. W. BEnnett.

In at least two gatherings of Desmids from the neighbourhood of Hindhead in Surrey, I have come across a phenomenon which I am not aware has been recorded before in this family of Algæ, viz. the formation of parthenospores without conjugation. The species was in all cases Closterium lanceolatum, Ktz., and I have at present seen four examples of it. In two of them one spore, in the other two two spores, were formed within the frond. They appeared spherical or ellipsoidal according to the view in which they were seen, and with perfectly smooth surface. The fronds were distinctly alive, and the longitudinal chlorophyll-bands had undergone but little change from their ordinary form, except where interrupted by the intervention of the spore. A similar phenomenon has been recorded in the allied Zygnemaceæ.

9. On a simple Apparatus for the Cultivation of small organisms in Hanging Drops, and in various Gases, under the Microscope. By Professor MARSHall Ward, F.R.S.

The author has found it necessary to devise a culture-chamber capable of supplying relatively large quantities of gases to the ordinary hanging-drop cultiva