In the composition of the bars the proportion of brass (the more dilatable metal used) is double that of steel, or two-thirds of brass to one of steel.

The maximum and minimum indicators are pieces of split brass, which, by its nip, causes enough friction to prevent their being moved by shaking or knocking, at the same time not enough to prevent the free action of the present-temperature indicator.

The pivot on which the indicator works is of gold, and the surrounding parts washed with the same metal, to prevent oxidation. This instrument has often been condemned from its want of agreement with the mercurial thermometers sent down with it; but now that there is proof of the effect on the latter by pressure, that disagreement is accounted for, and in most cases in which its notification has been rejected it would now be found correct.

Saxton's Metallic Thermometer.

In the Report of the Superintendent of the United States Coast Survey of 1860, is contained an account of the deep-sea thermometric instruments used in the survey of the Gulf-stream; and the instrument mainly depended on seems to have been an adaptation of the Breguet principle of the coil-form of two metals, by Joseph Saxton, Esq., and consists of two stout ribands of silver and platinum, carefully united with silver-solder and coiled spirally, the more expansive metal being within; a coil of gold, to which each metal is soldered, serves to prevent the tendency of the silver and platinum to separate: the lower part of the coil is affixed to a brass stem which passes up through the centre of the coil, while the upper end is fastened to the base of a short cylinder which keeps that end of the coil from turning, and consequently conveys the whole motion of the coil, caused by the unequal expansion or contraction of the two metals, to the lower end of the spindle, the turning of which is communicated to multiplying-wheels placed within the short cylinder, which serve to magnify the motion of the coil and move a hand, which in turn carries with it an index to record the extreme maximum or minimum.

This instrument was found to have a considerable error under pressure, and it is supposed that the solder by which the two metals are cemented together was so affected by the compression as to cause the error; but it may be due to the construction of the instrument itself, as there can be no certainty that the expansion and contraction of the two metals are communicated at the end of the spiral ribands, and may not be developed at the sides (the ribands not being close to the bar around which they are coiled).

If the communication of the contraction and expansion could be depended on at the end of the ribands, it would seem that a simple but long coil of wire could be so encased that the direct action could be communicated to an indicator for the purpose of registering the temperature indicated by its elongation or expansion and contraction.

DESCRIPTION OF WOODCUT.

Fig. 1. The instrument complete: one-sixth size.

Fig. 2. The stem and coil: full size.

a. Stem.

Fig. 3. Section: full size.

a. The stem.

b. Silver and platinum coil.

b. The coil. c. Short cylinder.

d. Magnifying wheel. e. Present-temperature indicator.

f. Registering indicator,

Fig. 4. Short cylinder: full size.

a. Magnifying wheel.
c. Registering indicator.

b. Present-temperature indicator,

Hearder's Metallic Thermometer

is the most simple and has the most direct action of the three: it consists of an iron tube, within which, and secured to the lower part by a socket, is a bar of another metal of unequal expansion with the iron; this unequal expansion is communicated by means of a short fulcrum to a long lever, the upper end of which is toothed to meet a small wheel correspondingly toothed to this wheel is attached the hand, the index being moved by the hand and kept in its place by means of friction.

An instrument of this description has not been completed as a thermometer; but the author has seen the action of an experimental instrument, and it seemed to answer well.

It is the invention of Dr. Jonathan Hearder, of Plymouth.

THE SMALLS LIGHTHOUSE.

THE present lighthouse on the Smalls Rock off Milford Haven is a noble structure, combining elegance with strength; the internal economy of the building is remarkable, while the illuminating apparatus is, if we may use the paradoxical expression, a scientific work of art. Much labour of subtle brains and lissom fingers has been expended in producing the rare mechanism and adjustment by means of which the present brilliant light streams out on to the adjacent sea. But this splendid lighthouse is, as many of our readers are aware, a comparatively modern structure, erected in 1861, in place of an old wooden building, which, although it had done good service in its time, in the latter years of its existence groaned and trembled with the weight of years and the violence of the waves.

We are enabled to publish a brief episode in the existence of the old building, the particulars of which contrast very strongly with the comparative efficiency and security of the present establishment as well as with the accommodation for the keepers.

The account is extracted from an unpublished memoir of an old sailor, whose life was a long series of dangers and hardships. The narrative is in some parts very graphic, and on its own merits is most interesting; it may also we think be serviceable in showing that if a hundred years shows such an advance in lighthouse science, we may still expect to progress, and as time goes on, to

attain to still greater perfection than we have now arrived at, so that the hidden dangers of the sea will be indicated in such a manner, that navigation will be robbed of half its perils and communication between all nations rendered more easy and general. We print the extract almost verbatim, save one or two alterations we have made to render it intelligible to our readers.

"In the summer of 1776 the lighthouse on the Smalls being finished, Mr. Philips, of Liverpool, the proprietor, called on me at my house at Hertfordwest, and gave me the offer of the government of the said lighthouse, after he had tried many others who refused it on account of the dangerous situation of the place. But I being willing to get a living for my family undertook it, and accordingly agreed with him for sixty pounds a year and my board, and after having entered into articles, embarked on October the third, 1766, with three others which were under my command at Solvah, and got on the rock the same day in the evening, and as soon as I arrived the engineer and his men returned by the same sloop. The weather continued favourable till about a fortnight, and on the seventeenth of October we were visited by a hard storm which washed several of the stays and braces that supported the pillars away; this storm continued two nights and two days and broke several of our lower room windows. Then I began to see that the building was by no means strong enough for the situation; however when the weather grew a little calm we went to work on the rock. One of the men being a smith, and having provided some bars of iron and a few tools, we made new braces and stays and secured them as well as we could. The building was an octagon standing upon nine pillars, three of them were iron, capped and jointed, and six wood; and they were let into the rock four feet and a half and fastened with lead. The length of the pillars from the rock was twenty-four feet to the bottom of the lower room of the lighthouse; the centre pillar was nineteen inches diameter, the three iron and the other five wood pillars were twelve inches in diameter. The stays which were all made of iron ran from the top of one pillar to the bottom of the other thus N, and were nailed to the wood pillars and cramped to the iron pillars, and after having secured them as well as we could we put our tools into the house which stood on the top of these pillars. There was a lower and upper story in the house fixed on the said nine pillars, the lower story was eight feet high and partitioned off, one part for oil, tools, etc., the other part was to sleep in. The upper story was the lighthouse which was twelve feet high, having eight windows which were large and glazed with thick glass, the panes were about twenty-two inches by sixteen