148 
Analyses of Books. 
which each of the observers must have conducted his labours. Borda’s deter, 
mination in 1792 was 39,12776, a result differing ,00067 inch from Biot's, and 
,00044 from Eater’s. The mean of Borda's and Biot’s is 39,1281, differing 
only njBits of an inch from Eater’s. 
In a postscript to this paper. Captain Sabine notices some trifling errors which 
had been pointed out. to him by Captain Eater in his former paper on the length 
of the seconds pendulum, ( Ph. Tr. 1827, Art. ix) ; he gives the amount of theerror, 
and a revised list of his determinations of the length of the pendulum in various 
latitudes, which is as follows : 
Eat. 
Length of P 
St. Thomas, 
0°. 24', 7 N 
39.02074 
Maranbam, 
2. 31, 6 S 
39.01213 
Ascension, 
7- 55. 2 S 
39.02410 
Sierra Leone, 
8. 29, 6 N 
39.01997 
Trinidad, 
10. 38, 9 N 
39 01888 
Bahia, 
12. 59, 3 S 
39.02433 
Jamaica, 
17- 56, 1 N 
39.03503 
New York, 
40. 42, 7 N 
39 10120 
London, 
51. 31, 1 N 
39.13929 
Prontheim, 
63. 26, 0 N 
39.17456 
Hammerfest, 
70. 40, 1 N 
39.19475 
Greenland, 
74. 32, 3 N 
3920335 
Spitzbergen, 
79. 49, 9 N 
39.21469 
V. On the measurement of high temperatures. By James Prinsep, Esq. Assay 
Master of the Mint at Benares. Communicated by Peter Mark Roget,iU. D. Sec. B-S. 
The author of this paper justly comments on the deductions in which the che- 
mists and natural philosophers of Europe appear to have hitherto reposed more 
confidence than they deserved. We mean the table of melting points, deduced from 
the indications of Wedgewood’s pyrometer, which continues to be gravelv copied 
from onesystem into another. We will venture to say, that noinquiringstudent ever 
received this table of results as other than purely imaginary, notwithstanding the 
high names that appear to have sanctioned it by adoption, and the reputation which 
the very ingenious author of the pyrometer so justly enjoyed. In England Mr 
Daniel, it appears, had the merit of first attracting attention to the subject, and he 
succeeded in showing, that the numbers till then received as expressing the degree of 
heat at which some of the metals entered inso a state of fusion, were so erroneous, 
as to prevent any confidence being placed in the indications of Wedgewood’s py- 
rometer. r ' 
Daniel's pyrometer, through a great improvement on the former, was still liable to 
some objections. Platina, the metal, the expansion of which is used to measure the 
degree of heat, has so low a rate of expansion that a very minute error in its esti- 
mated length would cause a considerable one in the deduced temperature. Add 
to which, that this little is diminished by the expansibility of the enclosed case of 
black lead. It may be doubted too, whether the expansibility of either the platina 
or black lead follows any law so regular, as to allow us to extend the scale, fixed 
by comparisons made at low temperatures, to very high ones. Indeed this ob- 
jection is not founded on mere conjecture ; for MM. Duloug and Petit have shown 
very satisfactorily, that no metal has the same rate of expansion at the higher 
temperatures as at the lower. It is further objected to Daniel’s pyrometer, that 
plumbago of which the case is formed is a very bad conductor of heat, and is also 
liable to lose its shape. 
There are two objects to which a pyrometer may he supposed to be applicable; 
that of indicating, by similar phenomena, equal degrees of temperature, and 
that of referring these temperatures to the more familiar and better known scale of 
Fahrenheit’s thermometer. The first constitutes the real value of the pyrome- 
ter ; the second is a question of curiosity. We think each of these problems has 
been resolved very happily by the author of the paper before us. 
To have a pyrometer that shall be universally comparable, that is that shall in 
similar temperatures develop similar phenomena j we have only to provide our- 
selves with small portions of metallic alloys of different degrees of fusibility, which 
being exposed in a cupel in any part of the furnace where we wish to know the tem- 
perature, the particular alloy that melts will indicate, so far, the value of the tern- 
perature ; that we shall be sure always to have the same temperature (within certain 
limits) where tins alloy melts. But we will allow the author to speak for himself. 
