396 
THE BRITISH ASSOCIATION MEETING AT 
EDINBURGH 
SECTION A. 
OWING to the large number of papers still remaining in 
Section A, the Section was divided on Tuesday into two | 
div'sioas. 
Thermo- Electricity, by Prof. Tait. The principal object of 
NATURE 







| 
| 
this research was to verify certain deductions which the author | 
had made from the principle of the Dissipation of Energy. But 
it had also a practical bearing, viz., to obtain a means of measur- 
ing high temperature, such as the melting point of iron or rocks. 
Sir W. Thomson’s great investigation had pointed out for the 
first time that the whole subject of thermo-electricity depended 
on something more than the ordinary and well-known Peltiet 
effect ; in fact, he predicted and subsequently verified by in- 
genious experiments the existence of what he called the specific 
heat of electricity, which, taken in conjunction with the 
Peltiet effect, supplied a complete explanation of the known 
phenomena consistent with the Conservation of Energy. His 
experiments further showed, if the relation between the electro- 
motive force of a thermo-element and the temperature be repre- 
sented by a curve in which the abscissz are proportional to the 
temperature and the ordinates to the electromotive force, that 
this curve is symmetrical about an ordinate. But they did not 
show the form of the curve. Some preliminary experiments 

made by the author of the present paper had shown the curve | 
to be so closely parabolic that he considered the subject worthy 
of a careful investigation. The difficulty of the research lay not 
so much in the experiments themselves as in obtainiyg wires of | 
the infusible metals. He tried circuits of almost every metal 
which could be obtained in suitable wires, platinum, palladium, 
iron, &c., and on plotting the results, all the curves appeared to | 
be parabolas. The form of experiment employed was this: As 
mercurial thermometers were inapplicable to temperatures 
above 300° C., the simultaneous indications of two separate 
thermo-electric circuits, the junction wires of which were im- | 
mersed in the same hot and cold baths, were plotted one as ab- 
scissa the other as ordinate. This method gives a very delicate 
test of the parabolic law ; for if the two curves obtained by plot- 
ting the two systems separately with the temperatures as abscis- 
sze be exact parabolas, then the curve obtained as above will 
also be an exact parabola ; but, if either of the former differ from 
the parabolic form, the latter will differ much more. All the 
results laid down by this method have as yet given very satisfac- 
tory approximations to parabolas. The consequences of this 
parabolic law are curious. If it be admitted, it proves the truth 
of the author’s deduction from the principle of the dissipation of 
energy, viz., that Thomson’s specific heat of electricity, like 
thermo and electric resistance, varies directly as the absolute 
temperature. The Peltiet effect is expressed as a parabolic 
function of the temperature. Another method of combining two 
thermo-electric circuits is to make the circuits, the junctions being 
in the same baths, act on a differential galvanometer in opposite 
directions. It is obvious, from the equations to the parabolas 
representing the relation of electromotive force to the temperature, 
or from considering the analogous case of the paths of two pro- | 
jectiles having the same horizontal velocity, and projected from 
the same point, that, by properly adjusting once for all tempera- 
tures the resistances of the two circuits, the term depending on 
the square of the temperature may be eliminated, and the galvano- 
meter indication will be proportional to the difference of tempera- 
tures of the two baths. 
Sir W. Thomson, speaking of the importance of Prof. Tait’s 
research, pointed out that its results were in direct contradiction 
to the statements made in books on the subject. 
A New Refiector for Lighthouses, by T. Stevenson. 
The novelty of this lantern, which was a holophotal apparatus 
with a spherical and approximately paraboloidal mirror, combined 
with a Fresnel’s lens, consisted in replacing the speculum which | 
usually forms the mirror by plate-glass facets, silvered, like a 
looking-glass, at the back. 
The paraboloidal mirror consisted of three annular facets formed 
by the revolution of a circular are about a horizontal axis, the 
circular are osculating the generating parabola to which it was 
required to approximate. 
The facets were prepared at Messrs. Chance’s works near Bir- 
mingham, by first bending plate glass into the approximate form, 
and then grinding and polishing the facets in the same way as the 
prisms of dioptric apparatus are ground and polished. The joints 

| be made osculating the surface, and be silvered at the back. Mr. 



between the facets should be made good with Canada balsam z ; 
they can then be hardly seen, as the refractive index of the bal- 
sam is nearly the same as that of the glass. 
It is sometimes necessary to construct a holophote which s 
illumine a given are of the horizon instead of the whole, as in th 
ordinary dioptric fixed light, or a very small portion alone, as i 
the revolving light; in fact, to send out a wedge-shaped beam 
diverging horizontally. This is usually effected by passing a por- 
tion of the light through a second system of prisms to deviate i 
into the required direction, after being rendered parallel in the 
usual way. But such a beam could be produced by a single re- 
flection from a suitable surface, and, though it would be impossible 
to construct such a surface in speculum metal, facets of glass could 







Stevenson had asked Prof. Tait for formule to calculate the form: 
of surface required. 
Prof. Tait exhibited the formule he had obtained, and pointed 
out that such questions could be easily solved by quaternions, and 
that this calculus was peculiarly adapted to solve the problems of 
geometrical optics. P 
A Method of Estimating the Distance of Fixed Stars, by Mr. Fox” 
Talbot. } 
The author did not know whether he had been anticipated in 
the proposal of the following method, and he left that to be 
determined by those better acquainted with practical astronomy. 
The principle of the method may be seen from a simple 
example. Suppose the plane of the orbit of a binary system to 
pass through the sun, 7.2, that the observer is in the plane of the 
orbit, and that in the spectra of the individual stars there are 
lines belonging to the same element. The spectra of the two 
stars taken through the same slit should be observed and com- 
pared. When the stars appear in the same straight line, it is 
clear that their velocities relative to the earth are the same, since 
bothare moving perpendicularly to the line of vision ; the lines 
from the two stars will therefore coincide. But when their 
apparent distance from each other is greatest, the difference of 
their velocities relative to the observer is equal to the velocity of 
either star in its orbit about the other. This difference of 
relative velocity will produce a displacement of the lines, which 
displacement may be observed and even measured. This will 
give us the value of that velocity. But we also know the periodic 
time. We have then at once the circumference and thence the 
diameter of the orbit. We know the greatest angular distance 
between the stars; we have then the distance of the stars from 
the earth. 
Repori of the Committee on Underground Temperature, by Prof. 
Everett, D.C. L. 
The intended boring at the bottom of Rosebridge Colliery has 
not been executed, recent occurrences in a neighbouring pit having 
given the manager reason to fear an irruption of water in the event 
of such a boring being made. Careful observations of tempera- 
ture have been taken by the engineers of the Alpine tunnel under 
Mont Frejus (commonly called the Mont Cenis tunnel). The 
highest temperature in the rocks excavated was found directly 
under the crest of the mountain, which is quite a mile overhead. 
The temperature was §5°1° Fahr., the mean annual temperature 
of the crest over it being estimated, from comparison with ob- 
served temperatures at both higher and lower levels (San Theo- 
dule and Turin), at 27°3° Fahr. Assuming this estimate to be 
correct, the increase of temperature downwards is at the rate of 
1° in 93 feet, which, by applying a conjectural correction for the 
convexity of the surface, is reduced to about 1° in Sr feet as the 
corresponding rate under a level surface. This is about the rate 
at Dukenfield Colliery, and is much slower than the average rate 
observed elsewhere. The rocks are extremely uniform, highly 
metamorphosed. and inclined at a steep angle. They contain 
silica as a very large ingredient. They are not faulted to any ex- 
tent, and are very free from water. It is proposed to sink two 
bores, to the depth of from 50 to 100 feet, at the summit, and 
another point of the surface over the tunnel, with the view of 
removing the uncertainty which at present exists as to the surface- 
temperature. Mr. G. J. Symons has repeated his observations 
at every fiftieth foot of depth in the water of the Kentish Town 
well, between the depths of 350 and 1,100 feet, the surface of 
the water being at the depth of about 210 feet. The observa- 
tions which have been repeated are thus completely free from the 
disturbing effect of seasonal changes. The results obtained agree 
closely with those previously found, and show between these 
depths a rate of 1° in 54 feet, which, from the estimated mean 
temperature of the surface of the ground, appears to be also very 
