Lord Charles Cavendish has also recorded a very accurate series of observations* on the depression of mercury in glass tubes, and these have furnished the basis not only for the correction of the reading of barometers, &c., but for the verification of the theory of capillary action by Young, Laplace, Poisson and Ivory.

I think it right to notice the scientific work of Lord Charles Cavendish, because Henry seems to have been living with him during the whole period of his electrical researches. Some of the jottings of his electrical calculations are on torn backs of letters, one of which is addressed,

[The Honble Mr Cavendish

at the Rt Hon ble

The La Charles

Cavendish's

Marlborough Street.

These calculations relate to the equivalent values of his trial plates when drawn out to different numbers of divisions. There is no date nor any part of the original letter.

The memoranda of some experiments similar to those in Art. 588, on the time of discharge of electricity through different bodies, are on the back of the usual Notice of the election of the Council and Officers of the Royal Society on the Thirtieth of November, 1774 (being St Andrew's Day) at Ten o'Clock in the Forenoon at the House of the Royal Society in Crane Court, Fleet Street. The address on the back of this letter is

To

The Hon Henry Cavendish

Gr Marlborough Street.

Dr Thomas Thomson, who was acquainted with Cavendish, says in his interesting sketch of him†,

"During his father's life-time he was kept in rather narrow circumstances, being allowed an annuity of £500 only, while his apartments

Phil. Trans., 1776, p. 382.

+ History of Chemistry, Vol. 1, p. 336, quoted in Wilson's Life of Cavendish, p. 159.

were a set of stables, fitted up for his accommodation. It was during "this period that he acquired those habits of economy and those singular "oddities of character which he exhibited ever after in so striking a manner."

The whole of the electric researches of which we are to give an account were made before the death of Lord Charles Cavendish, which took place in 1783. We must therefore suppose that they were made in Great Marlborough Street, and probably in the set of stables mentioned by Dr Thomson. He speaks of a "fore room and a back room" in Art. 469, and in Art. 335 he compares the size of the room in which he worked to that of a sphere 16 feet in diameter. The dimensions of his laboratory are of some importance in determining the electric capacity of bodies hung up in it, and by the foot-note to Art. 335 it would appear that the room was probably 14 feet high, which is somewhat lofty for "a set of stables," but I believe not much more than the height of some of the rooms in the dwelling-houses in Great Marlborough Street.

Let us then suppose that we have been admitted by Cavendish into his laboratory in Great Marlborough Street, as it was arranged for his electrical experiments in 1773, and let us make the best of an opportunity rarely, if ever, accorded to any scientific man of his own time, and examine the apparatus by which the electric fluid, instead of startling us with the brilliant phenomena, new instances of which were then every day being discovered, was made to submit itself, like everything else which entered that house, to be measured.

The largest piece of apparatus was the "machine for trying simple bodies" of which we have a description and sketch in Art. 241, and plans at Arts. 265 and 273. The framework of the machine is not represented in these figures.

We learn, however, from Dr Davy*, that

"Cavendish seemed to have in view, in construction, efficiency merely, without attention to appearance. Hard woods were never "used, excepting when required. Fir-wood (common deal) was that commonly employed."

* Wilson's Life, p. 178.

The bodies to be "tried" and the wires and vials for trying them were either supported on glass rods as shown in the sketch at Art. 239, or else hung by silk strings from a horizontal bar 7 feet 3 inches from the floor as mentioned in Art. 466. The electrical connexions were made and broken at the proper times by means of silk strings passing over pullies attached to the horizontal bar.

One of the bodies, the charges of which Cavendish compared by means of this apparatus, was a globe 12.1 inches in diameter covered with tinfoil. This globe has historical interest as it was not only the standard of capacity with which Cavendish compared that of all other bodies, but it formed part of the apparatus by which he established that the electric repulsion varies inversely as the square of the distance.

There was also a set of circles of tin plate, one of 36 inches diameter, one of 18.5 and two of 93; and also square and oblong tin plates, and squared pieces of stone and slate, and a collection of cylinders and wires of different sizes.

There was another "machine," represented, with its framework, in Fig. 20, Art. 295, “ for trying Leyden vials."

The "Leyden vials" were most of them flat plates of glass with circular coatings of tinfoil, one on each side. They were made in sets of three, any one of each set being nearly equal in capacity to the three of the former set taken together. Cavendish had thus a complete set of condensers of known capacity by means of which he measured the capacity of every piece of his apparatus, from the little wire which he used to connect his coated plates, and which he found to contain 28 "inches of electricity," up to his battery of 49 jars, which contained 321000 " inches of electricity" *.

These "inches of electricity" can be directly compared with our modern measurements of electrostatic capacity. Indeed the only difference is that Cavendish's " inches of electricity" express the diameter of the sphere of equivalent capacity, while the modern measurements express the capacity by stating the radius of the same sphere in centimetres.

* About half a microfarad.

Of each of these plates of glass Cavendish has given a most minute description, so that each, if it were found, could be identified. Mr Cottrell, of the Royal Institution, has been kind enough to examine the catalogue of apparatus there, which contains Cavendish's Eudiometer and Registering Thermometer. No trace, however, of a set of glass plates could be found. It is possible, however, that if the plates were neatly packed up, their small bulk and their apparent uselessness may have enabled them to survive the periodical overhaulings of some less celebrated repository, and that they may yet gain an honourable place in the museum of historical instruments.

But we need not expect ever to discover a piece of apparatus of still greater historical interest-that by which Cavendish proved that the law of electric repulsion could not differ from that of the inverse square by more than . It consisted of a pair of somewhat rickety wooden frames, to which two hemispheres of pasteboard were fastened by means of sticks of glass. By pulling a string these frames were made to open like a book, showing within the hemispheres the memorable globe of 121 inches diameter, supported on a glass stick as an axis. By pulling the string still more, the hemispheres were drawn quite away from the globe, and a pith ball electrometer was drawn up to the globe to test its "degree of electrification." A machine so bulky, so brittle, and so inelegant was not likely to last long, even in a lumber room. A facsimile of Cavendish's sketch of it is given at page 104. His own account of the experiment, in Arts. 217-234, is one of the most perfect examples of scientific exposition.

We might also notice the different electrometers, most of them consisting of a pair of cork or pith balls, mounted on straws or on linen threads, and some of them capable of having their weight altered by means of wires run into the straws; but though Cavendish had a wonderful power of making correct observations and getting accurate results with these somewhat clumsy instruments, we must confess that in these, the most vital organs of electric research, Cavendish showed less inventive genius than some of his contemporaries. When Lane and Henly brought out their respective electrometers, Cavendish compared their indications, and by

stating in every case the distance at which Lane's electrometer discharged, he has enabled us to calculate in modern units every degree of electrification that he made use of. What was really needed for Cavendish's experiments was a sensitive electrometer. Cavendish did the best with the electrometers he found in existence, but he did not invent a better one.

It was not till 1785 that Coulomb began to publish the wonderful series of experiments, in which he got such good results with the torsion electrometer, an instrument constructed on the same principle as that with which Cavendish afterwards measured the attraction of gravitation; and it was not till 1787 that Bennett described in the Philosophical Transactions the gold leaf electrometer, by means of which Volta afterwards demonstrated the different electrification of the different metals.

The electrical machine, by Nairne, was one with a glass globe. We should also notice the dividing engine, by Bird, for determining the thickness of the glass plates, and other small distances.

An attendant*, "Richard," appears occasionally, to help in turning the electrical machine, or in pulling the strings which made or broke the electrical connexions; and sometimes he is even asked his opinion as to the comparative strength of two electric shocks. But there is no record of any other person having being admitted into the laboratory during the series of experiments to which we now refer.

The authority of Cavendish in electrical science was of course established by his paper of 1771, and accordingly we find him nominated by the Royal Society as one of a committee appointed in 1772 "to consider of a method for securing the powder magazine at Purfleet 1."

A powder mill at Brescia having blown up in consequence of being struck by lightning, the Board of Ordnance applied to Mr Benjamin Wilson, F.R.S., who held the contract for the housepainting under the Board§, and who had some reputation as an

Arts. 242, 560, 565.

See Franklin's Works, Vol. v, p. 430, note.

+ Art. 511.

§ He also painted portraits of Franklin and of Gowin Knight, as well as of Garrick in various characters.