cated, by the nature of the dotted line, to have been irregular, or the near coincidence of the calculated and observed numbers in the table, and the variable sign of the differences, justify us in assuming the true maximum of vaporisation at the temperature corresponding to the highest point of the ellipse, namely, to 576° Fah.

At about 576° Fah. then, a bowl of copper *07 of an inch thick supplied with heat by a medium like oil, would be able so far to resist the cooling action of 60 grs. of water, as to produce the most rapid vaporisation; the quantity being sufficient to cover about onetenth of the surface exposed to heat.

5.25 inches, and the versed sine 1.45 inch; it deviated as little, therefore, from the figure of the last as could have been expected from the mode of forming it.

Nine observations were made of the vaporisation of one-eighth of an ounce of water in this bowl, placed in a bath of oil. Of these, seven are shown in the middle dotted line of fig. 1, plate 5, and agree very well with the ellipse traced in the full line; the two omitted were at temperatures lower than that of the lowest of the seven included in the figure. The following table shows the comparison of calculation and observation, assuming the major and minor axes of the ellipse to be respectively 251° and 214 seconds; and the co-ordinates of the centre 576° and 254 seconds. These values were not obtained rigidly, but they agreed better than numbers, greater and less, which were also tried.

No. of Observation.

foregoing, and the quantity of water used and nature of the bath were the same.

Of eight observations made and recorded in the following table, five only appear to belong to the same curve; this is seen in the lowest curve, plate 5, fig. 1, in which the dotted line represents the curve of observation. These five may be represented by a circle determined from observations 3, 4, and 8, which give for the radius 262°. The coordinates of the centre are 604° and 309 seconds.

Time of Vaporisation.

No. of Observation.

Fah.

168

Vaporisation in Iron Bowls.

13. Similar experiments were made with iron bowls of different thicknesses; No. V. 04 inch, No. II. 08 inch, No. VI. 18 inch, and No. III. of an intermediate thickness between Nos, H. and VI. The curvatures and general dimensions were intended to be those of the copper bowls, from which they in reality differed in no important particular. The radius of No. V. was 3.25 inches, of No. II. 3.1 inches, of No. VI. 2.9 inches; the chord of No. V. 5.2 inches, of No. II. 5.2 inches, of No. VI. 5.2 inches, the versed sine of No. V. 1.3 inches, of No. II; 145 inches, of No. VI. 16 inches. The difficulty of producing a uniform surface, and of retaining one of any smoothness, for a considerable time made these experiments much less satisfactory than those on the copper; in those with No. V. and No. II. oil obtained access to the cup and vitiated part of the results, and this was also the case at high temperatures with No. I. Small particles of water being thrown out of the dish, sunk below the oil without evaporating, and then in passing into vapour below the surface, threw up the oil with slight explosions. The surfaces were rough but clean, the quantity of water used oz. troy. The curves representing these observations are shown in plate 6; and through the striking irregularities in the three lower ones, we see the effect of thickness of metal in increasing the amount of vaporisation at a given temperature, the curve of No. III. being higher than that of No. II. and of No. II. higher than that of No. V.; and we also see a tendency towards a maximum lying above 540° Fah., though, from No. III. and No. V. obviously not far above it. The difficulty of passing the maximum with these thin bowls consisted chiefly in the acrid nature of the vapour given out by the oil, which acting powerfully on the eyes, rendered accuracy extremely difficult, and the effort sustained very painful.

With bowl No. VI. greater pains were taken to smooth the surface, and this was cleaned with alkali to free it from grease, and then with very dilute acid, which was washed off. The curve given to represent the observations is altogether more regular

In

If the vaporisation by the copper bowl No. VII. 07 inch thick, be compared with that of No. II. of iron 08 inch thick, it will be found to be much more considerable. fact, the curve traced for the copper bowl is exterior to the curve for No. III., and at the temperature of about 540° Fah. intersects that for the iron bowl No. VI. 18 inch thick. From 350° up to 508°, the time of vaporisation in the copper bowl varies from threefourths of that in the iron bowl of the same thickness, to three-eighths of the time, at corresponding temperatures. The specific heat of the iron being slightly higher than that of the copper, bulk for bulk, would tend to keep up the temperature of the former metal, but the conducting power of the copper being more than double that of the iron, would much more than compensate for its lower specific heat.

At

14. The effect of a surface covered with a thick coating or scale of oxide, may be seen by comparing the dotted line near the full line for bowl No. VI. with the full line. temperatures below 390° Fah. the scale of oxide diminishes the vaporisation considerably, probably by intercepting heat; but when repulsion begins to be developed, the scale acts to prevent it, and thus to raise the temperature of greatest vaporisation, and to diminish the time required for vaporisation at a given temperature. It will be recollected that this temperature of 390° differs but 7° from that found for the

vaporisation of drops from am

surface.

118 This circumstance will be recurred to again.

Quantities of fluid, varying from one-sixteenth up to one-fourth of an ounce troy, were now used with a view to ascertain the effect of varying the quantity upon the tem perature of maximum vaporisation. The surfaces were varied also. The results are given in the following table:

The effect of roughness of surface is to be seen in the three series; the effect at the lower temperatures seems to be generally to diminish the amount of vaporisation; and when repulsion would have taken place had the surface remained smooth to accelerate vaporisation at a given temperature, raising the point of greatest vaporisation on the scale. If this speculation be admitted, the temperature at which the rough and smooth surfaces vaporise equally, is but little above that of the real maximum of vaporisation of the metal when the cooling effect of the water is supposed to be entirely destroyed, that is, when the water is thrown upon it by small drops.

A comparison of the first and second series would place this point at about 386° Fah., the third and fourth at about 38810. The fifth and sixth would leave a doubt of its position, placing it by the nearest of two results at about 424°; while, on the other hand, the near approach at a lower temperature would incline us to make the coincidence conform more nearly to the numbers given by the other series, by selecting two less accordant times, at about 388° Fah.

The experiments on drops of water placed the temperature of maximum vaporisation in this same bowl at 334° Fah. when the surface was smooth, and at 346 when rough, no doubt a nearer approximation to the real point of maximum vaporisation than that. just deduced by the medium of a considerable quantity of fluid.

15. No satisfactory method occurred of ascertaining the temperature of a small portion of a piece of metal of the thickness used in steam-boilers, and exposed to the action of water, at or below the boiling point, while it received heat from a constant source. It was deemed advisable, therefore, to compare the effects which would be produced by commu nicating heat through a very good conductor, such as tin in the solid or liquid state, and through an imperfect conductor and cireulator, like the thickened oil employed in the foregoing series.

The same bowl was, therefore, tried in tin and in oil, with the same quantity of water, and with the following results, the bowl being 25 inch thick (No. VIII), and the material iron. The curves of observation are traced on fig. 2, plate 6.

Table of the Times of Vaporisation in different

Baths.

Tempera- Time in seconds.

ture.

One-eighth ounce.

Remarks.

Tin.

Oil.

The irregularity of the series made with the oil bath throws a doubt upon the maximum obtained, particularly as, with a thinner vessel, the preceding series gives a lower temperature as that of most rapid vaporisation, and the recurrence of the same time during a range of 19° confirms this doubt.

The temperature of greatest vaporisation in the tin was about 5083°, and the time but 6 seconds, while with the oil it was 9 seconds, as shown in this series, and probably less than 8, as shown in a foregoing series. The temperature of maximum vaporisation here given for the oil bath is 555°, differing 463 from that for the tin. Some where between 559° and 568° the times of vaporisation are the same for each bath, the repulsion due to the greater heat communicated by the tin counterbalancing the diminished vaporisation from the less heat given by the oil.

This comparison shows that the thickness of metal at which the effect of the material of the bath, or means of applying heat, would vanish, is by no means reached in practice.

16. With a less thickness of metal, this difference in the nature of the bath was, of course, more striking. In a dish, one-twelfth of an inch in thickness, the vaporisation, in a bath of tin, compared with a series made with the same surface, in an oil bath, was as follows: