Figure 2. Part of a bundle hanging loosely on the free under edge of the same section from which fig. 1 is drawn. The mosaic of primary bundles is unusually well marked. × 375.

Figure 3. Elastic fibre, with patches of chloride of gold deposit adherent. Isolated from adult human skin. Gold saturation and maceration in 20 per cent. acetic acid. x 340.

Figure 4. Isolated primary bundles.

Human adult skin. Gold saturation and

maceration in formic acid. × 340.

Figure 5. Tertiary bundle entwined by elastic fibres. Human adult skin. Gold saturation; maceration in acetic acid. × 340.

Figure 6. Gold deposit on a large elastic fibre, and a small elastic fibre on the surface of a bundle almost completely ensheathed in gold deposit. Human adult skin. Gold saturation; maceration in acetic acid. × 340.

Figure 7. An isolated secondary bundle, in which the contours of the primary bundles are visible. The latter are entwined by minute elastic fibres. Human adult skin. Gold saturation: maceration in formic acid. × 340.

Figure 8. Bundle showing fibrillæ, and snared by an elastic fibre (spiral fibre). Human adult skin. Saturation with glycerine; picro-carminate staining. (Hartnack, Objective No. 8; Eye-piece No. 3; Tube in.)

Figure 9. Lines of gold deposit on bundles, following the course of elastic fibres. Human adult skin. Gold saturation: maceration in acetic acid. × 340. Figure 10. Elastic fibre with a nucleus adhering to it, and a streak of gold deposit partially detached from the fibre. (Hartnack, Objective No. 8; Eyepiece No. 3.)

II. "On Hyaline Cartilage and deceptive appearances produced by Reagents, as observed in the examination of a Cartilaginous Tumour of the Lower Jaw." By GEORGE THIN, M.D. Communicated by Professor HUXLEY, Sec. R.S. Received November 25, 1878.

[PLATE 3.]

The following paper is written with a twofold object: firstly, as a contribution to the histology of hyaline cartilage; secondly, to illustrate how much the apparent structure of a tissue which is being examined microscopically depends on methods of preparation.

A portion of a large tumour of the lower jaw, believed from its naked eye appearances by two experienced surgeons to be sarcomatous in its nature, was given me for examination. Although I was struck by the peculiar kind of resistance it offered to the knife, I did not imagine at the time, any more than did the surgeons who excised it, that the tumour was cartilaginous. This is to be explained by the fact that the cartilaginous substance which had been growing with extreme rapidity was of a low type.

In order to determine the structure of the growth, I hardened por

tions of it in different solutions, and then made sections which I stained with various colouring agents. The sections thus prepared differed from each other in a remarkable manner.

Sections from a portion which had been placed for two days in solution of bichromate of potash were stained by logwood, picrocarminate of ammonia, and eosin respectively. In all of them the ground substance of the tumour appeared as structureless, and throughout it were interspersed a large number of rounded nuclei. In the carmine-stained preparations many of the nuclei were immediately surrounded by this homogeneous substance, without any appearances of what might have been considered as cell-substance or as cell-processes being observed. In some instances a scant, faintly granular stained substance tapered for a very short distance from opposite poles of the nucleus, producing the appearance of a spindle or fusiform cell. More rarely a long slender stained projection tapered gradually from one of the poles of the nucleus to a considerable distance, and seemed to end in a fine colourless fibre. The appearances were such as have been often described as indicative of cells with branching protoplasmic processes. For example, some of these apparent cells resembled accurately the smaller of the coloured figures described by Ranvier in the omentum as vaso-formative cells."

66

The sections stained in eosin solution showed somewhat the same appearances, although in a more exaggerated form. A homogeneous unstained ground substance was permeated by process-like prolongations of a finely granular stained substance which surrounded the nuclei, the prolongations from adjoining cellular centres anastomosing. The distribution of these cell-like masses of stained matter was an exact copy of the appearances seen in a cornea stained by gold chloride when what has been called the "positive image" is successfully produced, and would certainly quite recently, if not now, have been described by some histologists as a highly developed protoplasmic network of branching cells (fig. 14).

In the logwood sections the nuclei alone were stained, but stretching in various directions from the nucleus strong tapering colourless fibres appeared to be given off. A system of branched cells, in which the protoplasm was very scant, and the processes highly developed, was exactly simulated. This appearance was the more deceptive, as when the tissue was broken up with needles, numbers of these apparently branched cells with broken processes were found free in the fluid (fig. 13).

Slices of the tumour had been placed fresh in solution of purpurine (Ranvier's formula), and had been allowed to remain in it for several days. The surfaces of the slices were well stained, but the colouring action of the dye had not penetrated deeply. Thin sections were made from the stained surfaces and examined in glycerine.

The

nature of the tumour was at once apparent. Instead of the homogeneous ground substance seen in the other preparations, a typical hyaline cartilage, with a large proportion of so-called cartilage cells, was brought into view. The nuclei were well stained, but the cartilage substance proper was only very faintly coloured. In every part of all the purpurine sections the cartilage structure was perfect (fig. 11).

The purpurine solution contains one-third per cent. alum, and onefourth its bulk of methylated alcohol; and it is to this composition probably more than to the staining that the preservation of the unstable cartilage substance was due.

Portions of the tumour which had been placed in half per cent. solution of osmic acid were teased out, stained in logwood and examined in glycerine. Indications of the cartilaginous structure could be detected, but the preparations were chiefly valuable as demonstrating the nature of the cells. The cell-substance was stained a darkish brown colour, the nucleus was well stained by logwood, and the ground substance was very feebly stained. The outlines of the cells could thus be observed in situ, as well as studied in isolated cells, many of the latter floating free in the fluid.

All the cells observed were flattened, rounded, or somewhat polygonal bodies, with round nuclei (fig. 12). Their contours did not correspond exactly with those of the rounded cartilage "capsules" in which they lay.

In order to study the structure of these so-called "capsules," portions of the purpurine preparations were broken up. Considerable fragments with even surfaces were thus obtained, with rounded nuclei on the surfaces. In some of these there was no trace of the capsule formation. In other fragments a long piece of cartilaginous ground substance gave off laterally small curved projections, the size of the projection and degree of the curvature showing that they formed parts of a capsule. But in no instance was an entire capsule isolated. On the other hand, a curved projection could sometimes be traced round one side of a capsule, encircling nearly one-half of it, and then passing onwards to form the bent wall of another capsule. I never observed these projections doubling back round the capsule (fig. 15).

The examination of this tumour has thus shown that most delusive appearances as regards the nature of cartilage cells may be sometimes produced by staining and hardening agents. Carmine and eosin by staining an unformed substance that exists in the structure in defined tracts, may simulate branched protoplasmic cells, and bichromate and logwood preparations, either in sections or teased out, may as closely simulate cells with fibre processes.

These facts justify serious doubts as to the correctness of interpretation in all cases in which histologists have described branched cells

in hyaline cartilage, whether the latter existed as a normal structure, or as a pathological growth. They further show that, taken alone, carmine or eosin staining should not be held as conclusive evidence of the existence or limits of cellular protoplasm in any animal tissue.

EXPLANATION OF THE PLATE.

(All the figures are drawn by camera lucida; magnifying power × 260.)
HYALINE CARTILAGE.

Figure 11. The normal structure of hyaline cartilage. Purpurine.

Figure 12. Isolated cells. Osmic acid.

Figure 13. Isolated nuclei adherent to portions of cartilage substance, simulating branched cells with fibre-processes. Bichromate of potash; logwood. Figure 14. Stained substance in the cartilage simulating branched cellular protoplasm. Bichromate of potash; eosin.

Figure 15. Fragments of cartilage substance separated by needles. Purpurine.

III. "Volumetric Estimation of Sugar by an Ammoniated Cupric Test giving Reduction without Precipitation." By F. W. PAVY, M.D., F.R.S. Received December 5, 1878.

To be able to effect the quantitative determination of a body with accuracy and facility is an important matter looked at in relation to the study of its bearings. In the case of sugar there are no reliable means of precipitating and weighing it, either alone or in combination, and thus in the chemical estimation of this principle an indirect method has to be resorted to. The only property upon which dependence can be placed, for the purpose of chemical quantitative analysis, is its reducing action, under the influence of heat, upon certain metallic oxides, and that of copper is the one which general experience shows to answer best.

In the ordinary volumetric application of the copper test, the precipitation and diffusion of the reduced suboxide through the liquid interferes with the clear perception of the precise point of complete decoloration, and thus detracts from its delicacy. For purposes

where minute accuracy is of no moment, a sufficiently approximate result can be obtained, but for physiological investigation, and in other cases where precision is indispensable, the process is quite unfit for employment.

With the view of obtaining increased accuracy, chemists have had recourse to the plan of collecting the precipitate of reduced suboxide and weighing it as such or after reconversion into the oxide. From the difficulty, however, that exists in procuring the metallic oxide in a pure and uniform state, and from the impossibility of completely freeing the filter paper used from adhering surplus copper solution, some uncertainty is given to the results obtained by this method. To

obviate the difficulty here presented, I suggested, in a communication published in the "Proceedings of the Royal Society" for June, 1877, that the precipitated suboxide should be collected and dissolved, and the copper subsequently thrown down by the agency of galvanic action upon a platinum cylinder, as is now frequently done in the assaying of copper ores. The process has been found, as shown by the closeness observable in the results of counterpart analyses, to admit of the greatest precision, and I have turned it to extensive account in some recent physiological investigations I have conducted. In its application to such a purpose, it may be held that time and labour should be considered as of no moment, but it frequently happens that a more ready process of estimation is needed than the gravimetric supplies, and on this account a volumetric method, free from the objection I have pointed out as belonging to the ordinary plan, constitutes a desideratum.

A few years back Bernard introduced, for physiological purposes, a modification of the ordinary volumetric process, which is attended with reduction and the non-precipitation of the reduced oxide. The process involves the employment of a large quantity of caustic potash, and the presence in the product to be tested of extraneous organic matter. Under these circumstances it happens that the reduced suboxide is held in solution instead of being allowed to fall, and thus decoloration without precipitation occurs and enables the point of disappearance of the colour of the test to be ascertained with precision. Bernard, in his remarks upon the test, simply made mention of the fact that under these conditions, reduction without precipitation took place, but Dr. d'Arsonval,* his Préparateur at the College of France, refers the effect to the solvent influence of the extraneous organic matter in presence of the alkali.

Whilst engaged upon an inquiry into the merits of this test, the conclusion suggested itself to me that the agency preventing the deposition of the suboxide was the development of ammonia. With an absolutely pure solution of sugar, such as may be obtained by inverting the ordinary crystallized cane sugar (refined loaf sugar) no amount of potash will hinder the instantaneous precipitation of the suboxide. With commercial grape sugar, however, and in a still more marked manner with honey, interference with precipitation is temporarily exerted, and this, I am led to conclude, is due to the action of the potash in producing ammonia from the small quantity of nitrogenous organic matter incidentally present.

With this before me, the idea presented itself of resorting to the direct employment of ammonia for attaining the same result. It is well known to chemists that ammonia is a powerful solvent of the suboxide of copper, leading to the production of a perfectly colourless

"Gazette Hebdomadaire de Médecine et de Chirurgie," Sept. 14, 1877, p. 454.