TRANSACTIONS OF THE SECTIONS. 3 
_ difference is so small, that great heights are necessary for giving only a deviation of 
some tenth-parts of an inch. The experiments of Guglielmini gave indeed such a 
deviation; but at the same time they gave a deviation to the south, which was not in 
accordance with the mathematical calculations. De la Place objected to these expe-~ 
riments, that the author had not immediately verified his perpendicular, but only some 
months afterwards. In the beginning of this century, Dr. Benzenberg undertook new 
experiments at Hamburg from a height of about 240 feet. The book in which he de- 
scribes his experiments, contains in an appendix researches and illustrations upon the 
subject from Gauss and Olbers, to which several abstracts of older researches are added. 
The paper of Gauss is ill-printed, and therefore difficult to read; but the result is, 
that the experiments of Benzenberg should give a deviation of 3:95 French lines. 
The mean of his experiments gave 3-99; but they gave a still greater deviation to 
_ the south. Though the experiments here quoted seem to be satisfactory in point of 
re the eastern deviation, I cannot consider them to be so in truth; for it is but right to 
state that these experiments have considerable discrepancies among themselves, and 
i that their mean therefore cannot be of great value. In some other experiments made 
_ afterwards in a deep pit, Dr. Benzenberg obtained only the easterly deviation; but 
_ they seem not to.deserve more confidence. Greater faith is to be placed in the ex- 
periments tried by Professor Reich in a pit of 540 feet at Freiberg. Here the easterly 
a 
SIS EN 
deviation was also found in good agreement with the calculated result; but a con- 
i siderable southern deviation was observed. Iam not sure that I remember the num- 
_ bers obtained; but I must state that they were means of experiments which differed 
( much among themselves, though not in the same degree as those of Dr. Benzenberg. 
Professor Reich has published his researches, an abstract of which is to be found in Pog- 
gendorff’s ‘ Annalen der Physik.’ After all this there can be no doubt that our know- 
ledge upon this subject is imperfect, and that new experiments are to be desired; but 
these are so expensive, that it is not probable that they would be performed with all 
means necessary to their perfection without the concurrence of the British Associa- 
tion. I will here state the reasons which seem to recommend such an undertaking. 
1. The art of measurement has made great progress in these later times, and is here 
exercised in great perfection. 2. All kinds of workmanship can be obtained here in 
the highest perfection. I think it would not be impossible to have-an air-tight cylinder 
of some hundred feet high made for the purpose. This would indeed be expensive, 
but it would present the advantage that the experiments could be made in the vacuum 
and in different gases. 3. With these experiments others could be connected upon 
the celerity of the fall and the resistance opposed to it by the air and by gases. Pro- 
_ fessor Wheatstone’s method for measuring the time would here be of great use. 
_ 4. If the southern deviation should be confirmed, experiments could be undertaken in 
_ order to discover in how far this could be effected by magnetism in motion. For this 
purpose balls of different metals might be tried. Very moveable magnetical needles, 
_ well-sheltered, but placed sufficiently near to the path of the falling bodies, would 
_ indicate magnetical effects induced in them. 
On certain Cases of Elliptic Polarization of Light by Reflexion. 
By the Rev. Professor Powett, V.P.B.A. 
__ From the principle investigated by Fresnel, that polarized light changes its plane in 
_ reflexion, by a certain law dependent on the incidence, for transparent media, and the 
__ extension of a similar law to the reflexion from the second surface by Sir D. Brewster 
- (Phil. Trans. 1830), other formule were obtained by the last-named philosopher to 
_ express the varied phenomena observed by himself (Phil. Trans. 1841), in the re- 
_ flexion of polarized light from thin jilms, in extension of those previously investigated 
_ by Mr. Airy and M. Arago. The whole subject was reduced to the principles of the 
| undulatory theory by Dr. Lloyd (Brit. Assoc. 1841, Sect, Proc. p. 26), who pointed 
_ out the further theoretical result, that owing to the difference of phase or retardation, 
_ thus produced in the two portions into which the reflected light is divided, polarized 
_ light reflected by a thin plate will in general become elliptically polarized*. 
_ * This deduction, though stated in the report given in the Atheneum, is omitted in the 
volume of the Association, SR 
: BQ 
