tion of the remainder, while the other lead compound loses five equivalents of water at 200°.

The constitution of these substances may be viewed in a somewhat different light. Water, existing as such, combined with any substance allied to starch, either basic or constitutional, as supposed above, is never known to be so intimately combined with the substance as to be incapable of displacement by oxide of lead, especially under such a favourable circumstance as the presence of free ammonia. When substitution appears impossible, it is therefore reasonable to conclude that the water, if it exists at all, is neither basic nor constitutional, but is essential to the substance itself.

The remarkable circumstance of inulin giving off at one time three equivalents of water, and not at another, under apparently the same circumstances, suggests the idea that two inulines exist, having the same composition, but one containing water capable of displacement, which the other does not. This substance has, indeed, been obtained by chemists in two different physical states, gummy and pulverulent (although I have never obtained but the former dry); and I would suggest the probability of these having the constitutions above-mentioned, that is, that one is C24 H21 021, and the other C24 H18O18+ 3 HO; a feeble but unknown cause being sufficient to convert one into the other.

In

This is not a singular instance of a double constitution. In fact, the only satisfactory means of accounting for dimorphism, is by supposing a difference in the constitution of the same substance at one time to that which it possesses at another, the ultimate composition continuing the same. diabetic sugar, also, we have reason to suspect a double constitution, the descriptions of this substance by various authors being, as is well known, highly discordant. Some identify it with starch sugar; others, on the contrary, describe it as a peculiar substance. This is an interesting and very extensive inquiry, and would probably throw light on some of the cases of isomerism still unexplained. University College, London,

May 16, 1840.

XX. On certain Effects of Temperature. By C. Т. СолтHUPE, Esq.

To the Editors of the Philosophical Magazine and Journal. GENTLEMEN,

HAVING, from the nature of my occupations, an excellent laboratory for observing the effects of temperature, I beg

to offer you some experiments illustrative of some of these effects.

A modern glass-house is generally a cone built of brick, having its interior diameter at the base, varying from 58 to 60 feet, and its perpendicular height varying from 90 to 100 feet. The upper aperture through which the smoke ascends, varies from 9 feet 6 inches to 10 feet in diameter. This cone terminates at its base in substantial pillars of brick about 3 feet square, following the exterior inclination of the surface of the cone, and united above by arches which spring from pillar to pillar, and below by inverted arches beneath the ground.

Around the centre of the interior floor of this cone the furnace is erected; and around the exterior of the pillars which support the main body of the cone, the glass-house is extended by shed roofs, whose rafters bear against the exterior of the brick cone, above the arches which connect the pillars. This extension constitutes the manufacturing workshop, or space occupied by the glass-making operatives. The interior space around the furnace and within the pillars, is that occupied by the founders, or the men whose duty it is to fill the pots with raw materials for the production of glass, to urge the fire, to examine from time to time the state of fusion, and in short, to make from sand, alkali and lime, by the aid of intense heat, the material which the glass-making operatives subsequently convert by manipulation into glass.

For very many consecutive hours during the process of founding the raw materials, a thermometer placed at the greatest possible distance from the furnace, but within the area occupied by the founders, and freely suspended from a rod projecting from the interior surface of one of the brick pillars (a distance in the present instance = 20 feet 5 inches), will indicate a temperature varying from 316° to 325° of Fahrenheit. The founders have cool recesses, into which they frequently retire during their work, but the average of temperature here mentioned, viz. from 316° to 325°, and frequently very much beyond 325°, they bear without experiencing any inconvenience whatsoever. Strangers universally wonder at the possibility of human beings existing in a situation in which their clothes are continually scorched, while their naked skin exhibits no marks of the effects of fire. I had myself often wondered at the circumstance, until I made some experiments to endeavour to ascertain the cause of such an anomaly. The results of some of these experiments are curious from the extent of the ranges of the temperatures, and I have much pleasure in proffering them to those of my philosophical brethren who may feel an interest in such matters.

Exp. 1. Two silvered brass scale thermometers, having each a range of 600° Fahrenheit, were suspended from an iron pin at a distance of two inches from the interior surface of one of the brick pillars of the glass-house, at a distance of 20 feet 5 inches from the nearest point of the furnace emitting flame, and during the early part of the founding process.

They both indicated, from an average of ten simultaneous pairs of consecutive observations which ranged well together, a temperature of 194°4 Fahrenheit.

Exp. 2. One of the thermometer bulbs was now clothed with a thin case of fine black "merino." The average of ten simultaneous pairs of consecutive observations indicated a difference of temperature manifested by the clothed bulb, in excess of that manifested by the unclothed bulb, of 23°-1 Fahrenheit upon each pair of observations. Hence 23°-1 of Fahrenheit were retained by the covered bulb, which were evidently reflected, and lost to observation, by the bright metallic surface of the unclothed bulb.

In another series of experiments wherein the temperature indicated by the unclothed thermometers averaged only 1240.5 Fahrenheit, from twelve pairs of observations, the increment shown by covering one of the bulbs with a thin bag of black " merino," amounted to 340.66 Fahrenheit. Hence the quantity of heat that is reflected from the bright surface of a thermometer diminishes as the heat itself increases.

Exp. 3. During the latter part of the founding process and whilst the clothed thermometer suspended from the brick pillar ranged from 320° to 325° Fahrenheit, a small black iron cylindrical pan filled with water was placed upon a thin iron shelf, which had been fixed against the pillar and close by the side of the thermometer. It was reasonably anticipated that water thus placed in a temperature of 320° to 325° would boil; but after waiting until the half of it had evaporated, it showed no tendency to ebullition.

Exp. 4. The top of the iron pan was now covered with a pane of window-glass, and in a few minutes it boiled violently. This experiment demonstrated that the cooling properties of rapid evaporation can neutralize one of the direct effects of heat to a very surprising extent; and to ascertain the amount of this influence the following experiment was instituted.

Exp. 5. A clothed thermometer, whilst the mercury was indicating a temperature of 310° Fahrenheit, was immersed

in a vessel of boiling water. The mercury instantly fell to 212°, and then very gradually sunk to 141°. The merino envelope had become dry, and the mercury had commenced rising when the bulb was immersed a second time into the water. The mercury rose to 202°, and then gradually fell to 139°. By a third immersion the mercury rose to 198°, whence it fell gradatim to 133°. The envelope was now saturated with water at about 140° Fahrenheit, but the mercury speedily reassumed the temperature of 133o Fahrenheit, and remained at this fixed point for nearly five minutes, although the real temperature of its situation was, and had been for many previous hours, 310o Fahrenheit.

The effects of rapid aqueous evaporation were thus clearly shown to influence the indications of the thermometer when placed in a dry atmosphere of 310° Fahrenheit, and under the circumstances described, to the surprising extent of 177° Fahrenheit.

We may now infer that the copious perspiration which exudes from the skin of glass-makers, and of those who are engaged in similar scorching occupations, is a sufficient protection from the burning effects of a dry atmosphere of from 300 to 400 degrees of Fahrenheit; and that whilst the clothes of such persons are burning to tinder, their skin may be rendered insensible to the direct effects of fire upon the inanimate matter around them, by simple natural laws, viz. those of evaporation.

Having been engaged in some delicate experiments on the subject of heat, I was surprised at the effects of comparatively moderate dry temperatures upon such thermometer scales as were made of ivory.

In one instance the scale became shortened two degrees in 100 in a temperature of 212o Fahrenheit.

In another, an old and "well-seasoned" ivory scale that had often endured the maximum solar heat of Jamaica and the salt waters of the Atlantic, became shortened one degree in 110° from a dry temperature of 130° Fahrenheit.

In a third, the scale became shortened 11° in 120° from a short exposure to a dry heat of 260° Fahrenheit. Immersion in water will generally restore such scales to their original length.

I have the honour to be, Gentlemen,

Wraxall, near Bristol,
April 22, 1840.

yours obliged,

CHARLES THORNTON COATHUPE.

XXI. Remarks on some Tide Observations, published in the Transactions of the British Association. By RICHARD THOMAS, Civil Engineer*.

IN the Transactions of the British Association for the year

1838, is an account of a level line measured from the Bristol Channel to the English Channel, from which the writer draws the conclusion, "that the mean tide must be taken as the level of the sea," and that this mean tide is the same, or nearly the same elevation at all places, whatever may be the differences of rise and fall.

About three years ago the British Association published the results of certain tide observations made at Liverpool, which induced me to write the following letter on the subject, which was inserted in the 'Mining Journal,' and "Falmouth Packet" Newspapers.

" In the supplement to the Mining Journal,' No. 23, there appears to be published under the sanction of the British Association, 'that there is one invariable mean height common to neap and spring tides, the HALF TIDE MARK, a point from which engineers, geologists, and navigators will henceforward commence their calculations and adjust their standards of comparison.'

"I have reason to believe that however accurate the conclusion is with regard to the tides at Liverpool, where the observations were made, it is not correct as to its general application, and I mean to show that the tides generally have not the same elevation of half-tide mark, as applies to any particular locality, nor is the average half-tide mark, nor low-water mark, nor high-water mark at one part of the coast to be depended on as level with the corresponding tide marks on other parts. More than twenty years ago I had occasion to attend particularly to the tides at Falmouth, and the result of my observations showed differences as much as two feet and a half in elevation on half-tide marks. The rise and fall of ordinary spring tides there is about 17 feet, and of ordinary neaps little more than seven feet, but the several rounds of tides differ as to the mean elevation of the sea; the low- and highwater marks for the same difference of rise and fall being at greater elevations than others.

"At King-Road near Bristol, I observed the tides in the year 1815, and found that the difference of elevations of some half-tide marks amounted to about four feet.

" It may be possible, and I think it probable, from my observations, that these differences of half-tide levels, or rather