Fune 10, 1886] 
Everett ; and delete the word ‘‘ Thus” in the sentence which, 
at present, (wrongly) follows instead of preceding it. This 
change is obviously called for by the context :—for the reader 
has just been told how far theory had guided Thomson as to 
certain ‘‘ absorptions,” &c., of heat ; and, of course, expects 
next to be told what additional information, as to these 
‘‘absorptions,” &c., Thomson obtained by experiment. 
Still, confused as it is, the passage could not (except possibly 
from the point of view of history) embarrass a reader of § 196; 
for the nature of the Thomson effect is there ayazn clearly 
stated, and even illustrated by a diagram. [A much more 
serious case of confusion is to be found at p. 366, line 15; 
where (by the omission of a few words) my copyist has made 
absolute nonsense of a quotation from Clerk-Maxwell. ] 
The statement quoted by Dr. Everett obviously requires to be 
restricted, as follows :— 
An electric current, passing from cold to hot in copper, behaves 
as a real fluid would do :—i.e. it tends to reduce the gradient 
of temperature. In iron, under the same circumstances, it tends 
to increase the gradient. 
It is clear that this statement has nothing to do with the 
general nature of the Thomson effect :—.e. ‘‘ absorption” or 
“disengagement ” of heat :—for ¢4zs would depend upon the 
temperature of the fluid spoken of. It raises the question of the 
excess of Thomson effect in one locality, over that in another, at 
a lower mean temperature but with an equal gradient. 
Dr. Everett seems to forget that, though the water-equiva- 
lent of a metal may be treated as sensibly constant through 
moderate ranges of temperature, the j* specific heat of elec- 
tricity” cannot so be treated. Using nis notation, (with the 
proviso that @ is aéso/ute temperature) we have o = £0, and the 
equation he quotes from Thomson is 
a _ _ 0 dd 
dt c ax 
Happily, this can be integrated, so that we have 
o=F(x- 40). Aes hae yan (1) 
Now suppose the gradient of temperature to be uniform and 
positive along x positive (the direction of the unit current) ; 
when ¢= 0 we have 
(eS are 
Generally, therefore, 
Thus the gradient becomes less steep :—7.e. there is a tendency 
to reduce temperature differences, when & is positive, as in 
copper. In iron, where & is negative, the tendency is to make 
the gradient steeper :—7.¢. to exaggerate differences of tempera- 
ture. Of course, as in all these thermo-electric matters, reversal 
of sign of the gradient reverses the thermal effect. 
The general integral (1) denotes a process of continued simle 
shearing, not translation, of the “‘temperature curve.” Were it 
not for heat-conduction, harmonic waves of temperature would 
tend to become Jreakers, But it is idle to speculate farther. 
How much of this is Thomson’s I don’t certainly know ; and I 
_ am for the present too busy to enquire. But it would be difficult 
to overestimate his services to Thermo-electricity. 
_ This will, I hope, meet with Dr. Everett’s approval. As to 
his letter, I would say (in Scottish legal phrase) ‘‘ Quoad ultra, 
denied.” PiG. DA 
May 28 
Power in Laboratories 
IN connection with the admirable devices for the distribution 
of driving-power in laboratories, illustrated in NATURE, vol. 
xxxill. p. 248, the description of a novel and very effective form 
of water-engine, with which I have been experimenting for 
several months, will be of interest. 
One of these motors is set up in the cellar of our science hall, 
where it is supplied with aqueduct-pressure of sixty pounds to 
the square inch, and the power is transmitted from it by means 
of rubber belting led over ‘‘idle pulleys” to the upper stories 
of the building, where a small engine-lathe and dynamo are 
NATURE 
1 
driven. A word will suffice to explain the very simple construc- 
tion of the motor—a system of radial cylinders, with their bases 
at the centre of the motor, through which runs the driving-shaft. 
The pistons in these cylinders are single-acting, and the water is 
admitted to them in succession by the rotary valve which forms 
part of the main shaft. The pistons, thus, in pressing outward, 
exert their force against a strong ring, to which is bolted a cross- 
bar which engages the crank of the main shaft. Thus the ring, 
in turning the shaft, has the vibratory motion of an eccentric, 
and returns the opposite pistons to the bases of the cylinders, at 
the same time exhausting the water through the interior of the 
rotary valve. Three pistons are thus constantly exerting a thrust 
upon the ring, whatever its position, and this thrust being always 
tangential to the arc of revolution of the crank, there is no 
‘‘ dead centre,” and the uniform pressure at right angles to the 
crank at every part of its arc insures an even rotary motion and 
obviates the necessity of a balance-wheel. The ends of the 
piston-rods are slotted, and contain anti-friction rollers which 
bear against the ring, and this latter is grooved all round, so 
that, in addition to its simple and rapid motion as an eccentric, 
the ring is free to perform a slow motion of revolution indepen- 
dently of its work of driving the crank, and the wear of the 
anferior face of the ring is thus equalised and becomes inappre- 
ciable. 
The supply-pipe for this motor has a diameter of 14 inches, 
and it gives an equivalent of nearly 2 horse-power. The flow 
of water is regulated by means of a balanced valve, under control 
from every point where the power is used. As the use of the 
power is, for the most part, discontinuous, like that in lathe- 
work, I find it better to start and stop the motor as often as 
desired than to use the ordinary device of shifting a belt off and 
on a loose pulley. All possible economy of water is assured, as 
Side View. Front View. 
none of it runs to waste without giving its equivalent of power 
at just the time when itis required. It will be seen that this 
form of motor is specially adapted to such uses, as there is no 
fly-wheel whose inertia has to be overcome; and as the motor 
has no ‘‘dead centre,” it readily starts from any position, over- 
coming a maximum resistance. 
Where continuous running is required, at an invariable speed, 
a centrifugal governor is attached to the belt-wheel, and acts 
upon the amplitude of vibration of the ring, diminishing the 
stroke of the pistons when the resistance is removed. The 
governor thus gauges the water-supply exactly proportional to 
the resistance to be overcome, and makes the motor a very 
effective driving-power for dynamos and all sorts of machines 
and apparatus in which a uniform speed is necessary, while the 
resistance is variable. 
The difficulties barring the economic use of water as a motive- 
power, owing to its weight and incompressibility, seem to have 
been successfully overcome in this form of motor, with which 
unexampled speeds have been attained, and more than 80 per 
cent. of the theoretical power of the water derived. The little 
cut annexed shows the smallest size of these motors—it stands 
about 10 inches high, and uses a }-inch supply, consuming less 
than six quarts of water in 100 revolutions. I frequently run it 
at a speed of 1000 revolutions to the minute, and at the manu- 
factory I have seen the same motor attain double this velocity. 
The motor runs equally well with compressed air (or with 
steam, if the piston-packings are changed), and with either of 
these media even higher speeds are attainable. 
I find that the constant readiness of the motor for the imme- 
diate development of power, the little care it has required (only 
occasional oiling), and its economical consumption of water, are 
