and anomalous expansion. There is at that point a knot of great complexity and significance. When it is unravelled, there may te something decisive bearing on the present discussion.

Monday, 5th July 1880.

MR ROBERT GRAY in the Chair.

The following Communications were read:

1. On Peroxides of Zinc, Cadmium, Magnesium, and Aluminium. By J. Gibson, Ph.D., and R. M. Morison, D.Sc.

2. On the Processes in Subepiphysal Bone Growth and some points in Bone Resorption. By De Burgh Birch, M.B., Demonstrator of Physiology in the University of Edinburgh.

(Abstract.)

Subepiphysal Bone Growth.-Two processes must be noticed in this connection.

1st, The replacement of the neck of the cartilaginous head or epiphysis by cancellous tissue as an accompaniment to the rise of the epiphysis caused by the growth of the cartilage forming its neck. The cartilage is channelled by the advancing marrow, the rows of cartilage capsules being opened up.

The opening up of the rows of cartilage capsules results from the presence of a capillary blood-vessel forming the head of the column of marrow which lies in immediate contact with the next unopened capsule (Ranvier). The close proximity into which the pabulum is thus brought with the cartilage corpuscle in the unopened capsule nearest it causes it to grow rapidly and absorb the surrounding cartilage, this occurring, quickest in the direction of least resistance that is, towards the marrow.

The cartilage capsules communicate with each other by means of fine channels, a fact already hinted at by Budge.

The osseous tissue which is deposited upon the cartilage spicules, or septa which result from the channelling of the cartilage, forms the cancellous tissue; this forms a stable base off which the epiphysis

A

rises by growth of the cartilaginous zone immediately above the primary cancellous spaces.

2nd, The extension of the shaft occurs by opposition to its extremity, thus keeping up with the recession of the epiphysis. The extension of the shaft in length does not lift the head.

The area of proliferation in which these changes occur at the end of the shaft lies in an angular groove at the point where the neck joins the epiphysis, called by Ranvier encoche d'ossification.

This author describes fibres in the outer part of the encoche, that is the periosteum, which stretched from the periosteum to the cartilaginous head in which they became lost.

The existence of these is undoubted, and very general.

Origin of the Osteoblasts.-These organisms, which have the function of resorping bone, occur in certain well-marked situations; their origin is from the perivascular connective tissue, i.e., within a short distance of the line of ossification under the epiphysis, and extending over a considerable area, they diminish the number of spicules opening up the cancelli. Under the head of bones, the epiphysis of which have a greater sectional erea than the shaft, and in those positions where the head projects beyond the shaft externally along the interior of the shaft wall.

3. On the Wire Telephone and its Application to the Study of the Properties of strongly Magnetic Metals. By Professor Chrystal.

Four distinct sources of sound were noticed in the course of the experiments.

1. The variation of the longitudinal tension of the wire, owing to variation in the heating, still appears to be the most likely explanation of the action of the wire telephone, when a very fine wire of ordinary metal is used. Experiments were tried with induction coils of various sizes, the violin and microphone being put into the primary circuit and the fine wire telephone into the secondary. It was found that the sound diminished as the spark-giving power of the coil increased. With Professor Tait's large induction coil no sound at all could be obtained, when the secondary was closed through the most sensitive wire I possess.

2. It was found, however, that when the secondary circuit was broken, loud sounds were emitted at the pools of the mercury break. These sounds appear to be due to electrostatic action. They are most probably of the same nature as those obtained in Thomson's singing condenser, Edison's condenser telephone, &c.

3. If the wire of the wire telephone be placed across the lines of force in a strong magnetic field, very loud and pure sounds are obtained when a current interrupted by a tuning-fork is passed through it. These sounds can be obtained with very thick wires of any metal. If a tolerably thin wire be used, although the sound is not much louder, the amplitude of the vibration increases; as much as 2 mm. was observed.

4. Experiments were also made with a view to explain the anomalous behaviour of iron wires established by De la Rive and Dr Ferguson.

Experiments with soft iron wires showed that the sounds did not, in the case of iron, depend in the same way on the length and thickness as they do in the case of ordinary metals, and that their quality is essentially different. The note of the interruptor is often not heard at all, but instead, a variety of other notes are produced, some of them very high accompanied with a fizzing or buzzing noise.

The sound depends on the temperature of the wire, being loudest about a dull red heat, just above the temperature at which the abnormal extension and contraction and the re-glow are usually observed. At higher temperatures the sound falls off very rapidly.

These results suggested that the sound is a consequence of the magnetism of the iron; for, in the case of soft iron, the magnetic susceptibility is at a maximum about the temperature above mentioned, and falls off very rapidly at higher temperatures.

Experiments with steel wires settled the question, for it was found that when the steel was made white hot and then tempered, so as to deprive it of its permanent magnetism and make it hard, it gave no sound at all in the wire telephone. On magnetising it, however, by stroking once or twice with a bar magnet, it sounded quite distinctly, giving a high note and a soft fizzing sound.

The effect of heating a magnetised steel wire is as follows:-At first the sound falls off, first the fizzing disappears, then the high note; then comes an interval of silence; then, as the temperature

increases, the high note comes in again; then the fizzing sound, which quickly rises to a deep buzz accompanied by several notes, among which may be heard the note of the interrupting tuning fork; as the temperature goes on increasing, these sounds die out again in the corresponding order, and when the whole wire is bright red, absolutely nothing can be heard.

All these effects are explained by the magnetism of the steel. The first effect of heat is to destroy the permanent magnetism, which about 250° C. is practically insensible; above this temperature the susceptibility for induced magnetism increases very fast, reaches a maximum about dull red, and then falls off again.

Advantage was taken of Professor Tait's thermoelectric diagram to verify the close connection between the magnetic, thermoelectric, and other characteristic physical properties of iron and its power of producing sounds, when traversed by a varying current of electricity. The agreement was found to be very striking.

Similar experiments were made with nickel, which is remarkable for the low temperature at which it loses its magnetic susceptibility. The behaviour of a nickel strip in the wire telephone was exactly in accordance with its magnetic properties. The results of thermometric and thermoelectric measurements, rendered the agreement still more remarkable.

Cobalt, when magnetised and heated, gave first a minimum of sound and then an increase; but no maximum was reached at the highest temperature (a bright red), to which I exposed it. This, again, is what is to be expected from its magnetic properties.

Both with cobalt, and with steel which had been softened by heating to a high temperature, the effects due to permanent and to induced magnetism interfere, so that no period of absolute silence appears. Occasionally this interference produces very strong beats. A full account of the experiments above alluded to will be published in "Nature" (vol. xxii. No. 561, p. 303-July 29, 1880).

In the thermo-electric measurement above referred to I had the able assistance of Dr Knott, whose experience in such work is well known to the Society. The curves from which the above references were drawn were constructed by him, and will be given in the detailed account of the experiment to be published in "Nature."

4. Notice of the Completion of the new Rock Thermometers at the Royal Observatory, Edinburgh, and what they are for. By Professor Piazzi Smyth, F.R.S.E.

The nature of this paper may be understood from the following headings:

(1.) The making and placing of the new thermometers. (2.) Practically described by Mr Thomas Wedderburn. (3.) The problem with the old thermometers.

(4.) Their next use in level fluctuations.

(5.) Their employment by Sir Wm. Thomson.

(6.) Their subsequent demonstration of the cycle of supra-annual waves of heat and cold.

(7.) The published predictions in 1872 for 1878-80.

(8.) The spoiled predictions in 1877, under the influence of erroneous sun-spot dates.

(9.) The rectified predictions in 1879, when the true date of sunspot minimum was ascertained by direct observation.

(10.) How to obtain correct dates for future sun-spot minima. Appendix I.-The contract for the new thermometers. Appendix II.-Account of works by Mr Richard Adie.

Appendix III. Further account by Mr T. Wedderburn, of Adie & Son.

Appendix IV. The cyclical seasons of 1878-80, as predicted in 1872.

Appendix V.-Scottish meteorological data from 1821-1880, arranged in quadruple annual means for cyclical inquiries. Plate representing the above numerical tables, graphically.

Monday, 19th July 1880.

PROFESSOR SIR WYVILLE THOMSON, Vice-President, in the Chair.

The following Communications were read :

1. Report on Fossil Fishes collected by the Geological Survey of Scotland in Roxburghshire and Dumfriesshire. Part L -Ganoidei.

By Dr R. H. Traquair.