JOURNAL ARY

OF THE

ROYAL MICROSCOPICAL SOCIETY.

AUGUST 1884.

TRANSACTIONS OF THE SOCIETY.

XI.-Researches on the Structure of the Cell-walls of Diatoms.* By Dr. J. H. L. FLÖGEL.

(Read 12th December, 1883.)

PLATES VIII. AND IX.

THE various objections made to my previous researches have induced me to undertake a re-investigation, and although this has for the most part been simply corroborative, it has appeared to me desirable to publish the results, together with some obtained in reference to other diatoms of which only preliminary communications (7) have hitherto been published.

Day by day the old erroneous views on the structure of diatoms are repeated in the text-books of Botany, and nobody seems to have approached the subject seriously; the present paper may therefore, I hope, arouse sufficient interest to induce further investigation.

I. METHOD OF INVESTIGATION.

I have already (6) described the methods which I have followed in the investigation of Pleurosigma; but it is necessary to refer here to the method of section-cutting. In applying my greater experience to Pleurosigma and other diatoms I have obtained generally much better results than formerly, which have lent additional confirmation to my paper of 1870.

Time of making Sections.-Accidentally I made my original investigations during the summer. When in the winter I wanted to replace some preparations, I found it impossible to obtain successful results. The cause was the artificially heated room. It is useless to attempt to bring the gum to the right degree of hardness by the addition of glycerin, sugar, &c., as, in consequence of the *The original paper is written in German, and has been translated by Mr. J. Mayall, jun.

Ser. 2.-VOL. IV.

263319

2 M

AUMULIOO

proximity of the face and the hands, the aqueous contents of such strongly hygroscopic substances are subject to an uncontrollable change, so that we must work with pure gum arabic and in the

summer.

Slight differences are caused by the weather or by the position of the sun. Sections of the coarser diatoms which are not required to be of extreme thinness are best made with a clouded sky or in rainy weather. Sections of Pleurosigma succeed best when the sun shines directly into the laboratory without striking upon the Microscope. I can hardly too strongly dwell upon the necessity of entering upon such experimental investigations under the best physical conditions, absence of vibration, noise, &c.

Placing the Frustule.-By my earlier method sections in different directions were obtained hap-hazard because the frustules were lying pell-mell in the gum. I have now improved the method as follows:-I take a number of frustules like a bundle of rods and cut sections in the exact transverse or longitudinal direction. With this important but difficult method every one must become familiar if he intends to check the results (hereafter described) which I obtained with Pinnularia.

(1) If we wish to examine uninjured specimens, the diatoms are first stained, usually by picro-carmine; they are then put in absolute alcohol. A glass slip is coated with collodion, and allowed to set. To avoid peeling when dry, the collodion must not be too thick. A drop of thick gum is then put on. A cluster of diatoms is taken direct from the alcohol with forceps and placed in the gum. In consequence of the current set up, the diatoms immediately distribute themselves equally through the gum. As soon as the edge of the gum begins to harden, one frustule after the other is drawn to the edge by a very fine needle, where, with proper manipulation, they can be piled up like a bundle of rods. All those which interfere with this piling up should be removed. Owing to the staining, the frustules can be readily seen on the transparent ground. As soon as the edge dries, a drop of fluid gum is added by a needle, and this process is repeated until the solid layer of gum has reached such a thickness that a displacement of the frustules need be no longer feared; a small patch of collodion is then put on. The preparation is now cut out by four cross-cuts and carefully removed from the glass; the bundle of diatoms in gum being contained between two films of collodion. It is advisable to make a drawing with a high power to show the position of the individual frustules and aid in the identification of the sections. The preparation is then put upon a nearly dry flat drop of gum on a piece of cardboard, and the base and the edges are made to adhere, if necessary, by a few drops of water, and by the addition of minute drops of gum it is so imbedded that at last it is entirely surrounded. This must be carried

out so cautiously that the collodion films do not separate; very careful watching of the imbedding process is therefore necessary, especially also to avoid cracks which commence at the edge and might easily extend to the object. From this bundle, sections may be made according to the method described by me in 1870.

(2) Any one wishing to study the structure of the individual valves and mark their manner of combination can shorten this somewhat detailed process. A cluster of diatom valves is taken out of the alcohol and placed in a large drop of water on a slide; and the water is allowed to evaporate after the valves have been evenly distributed. A drop of gum already dry-if possible with a flat surface on a piece of cardboard-should be in readiness; another piece of glass is coated with oil of turpentine, which is allowed to run off so that a very thin film is left, which does not readily dry. In this the point of a fine needle is dipped vertically, taking up sufficient oil so that by touching a frustule lying on another slide it will adhere. Thus the frustule can be put on the hardened drop of gum which has been moistened by the breath; this is repeated with a number of valves ad libitum, and finally they are covered with minute drops of gum till the required thickness is attained. This transference of dry frustules upon the dry gum is much easier than the process with fluid gum described under (1), because, with the latter, it frequently occurs that in bringing a new frustule into place the others are disturbed. With uninjured frustules process (2) is not available, because these, after the drying of the thin upper layer of gum moistened with the breath, will at once become charged with air and baffle any cutting. This absorption of air can only be avoided by transferring the frustules from the alcohol directly into the fluid gum, which then diffuses equally through them. At the most the frustule at the moment of hardening is slightly compressed, which injures somewhat the appearance of the sections.

Making Sections.-Numerous attempts to cut diatoms on the microtome failed, and I always returned to cutting by hand, under a dissecting Microscope. Gum is not favourable as the imbedding medium for the microtome. If a better medium were discovered (paraffin is useless), then a new era would open for these researches. Knives with broad backs should be employed; the angles of inclination to the cutting edge I have used are 21° 20'.

Piling-up of the Sections.-Pfitzer (19, p. 42) formerly proposed to moisten by the breath the gum-chips containing the diatom sections after they had been put upon the slide, whereby they naturally adhered. Anything more unpractical cannot be imagined; the very thing one wishes to avoid-namely, the disturbance of the sections-is by his method certain to occur, and in the most favourable case we have a hardened drop of gum in

which are scattered all sorts of fragments of the diatom sections, which the observer may be able to define with reference to their previous position, but which are utterly useless for the study of real details of structure. The unavoidable proximity of finger and face during the cutting is injurious, because the immediate surrounding atmosphere of the operator always contains a quantity of moisture which causes the delicate chips of gum immediately to adhere. In piling-up the sections we should most carefully avoid every increase of moisture, and during the operation the breathing should be suppressed. If after discharging the sections from the knife we do not intend to put on the cover-glass at once, then the slide with the chips should only be lifted up when in sunlight and under protection from dust. As a rule, time should not here be lost. It is advisable before putting on the cover-glass to touch two corners, which first come in contact with the slide, with small drops of balsam, so that during the lowering of the cover-glass it does not slip out of place and so grind up the very brittle chips of gum. The lowering of the cover-glass should be done slowly and steadily. If let fall, the puff of air will blow away the chips. If the chips, on account of too little moisture in the atmosphere, are curled up, the cover-glass will flatten them out, the gum on the edges may split, but the centre becomes flat and is often very useful. In general, however, it is recommended, on the days when these difficulties occur, to postpone the operation. Once for all, I state that flat sections are always the most instructive. After putting on the cover-glass a proportionately small drop of thin fluid-balsam is added on the edge-only so much that the sections are imbedded in it. After a few days the vacant spaces are filled up by fresh drops of balsam; a derangement of the previously filled-up sections is no longer to be feared.

Series Preparations.-With Pinnularia the making of series sections is almost a necessity, especially for the longitudinal sections. My procedure is as follows, though I am well aware it is capable of improvement. I cut off No. 1, leaving it on the edge of the razor; then I take a second section a little further off, and so on, until five are on the edge." These I transfer to the slide with a needle, and as nearly as possible in a straight line. This is repeated with the other five sections which form the second line. Then I take a new slide, and thus get decades of sections from a bundle previously prepared. A disadvantage is that sometimes the sections do not remain on the edge; by falling upon the table they are lost. Also that not seldom one cuts a gum-chip as a numbered section, which on further examination proves never to have touched the bundle of frustules. By disregarding these very troublesome mishaps, the series-section method according to my view gives the most beautiful results obtainable with such