412 
NATURE 
fAucust 28, 1902 
antiquity. He proposes a unit the “cé” or 1/1ooth of the | 
day, subdividing it into the “ décicé” and “millicé.” In the 
meantime, he would confine the system to men of science, 
but would teach it in the schools as soon as it meets with 
international approval. M. Goedseels considers the 
greatest obstacle to progress to be the existence of 
numerous tables and costly instruments based on the 
sexagesimal system. To help to remove this obstacle, he 
contributes seven pages of tables for converting time and 
angles to a decimal system. He takes the hour and the 
degree as units for one system ; fora second he supposes 
the day divided into forty hours, the circumference into 
400 grades. Dr. F. Jaja advocates a system similar to 
that of M. de Rey-Pailhade ; but instead of “cé” he calls 
his unit “degré,” its multiples ‘“décagrade,” ‘“hecto- 
grade,” its submultiples “ décigrade,” &c., down to 
“ décimilligrade.” For use by the public, he suggests 
for the subdivisions the titles ‘‘ minute premiére,” “ minute 
seconde,” “ moment” and “ instant.” 
The English equivalent of a “ minute seconde” would 
be found rather awkward, and why should an “ instant ” 
be shorter than a “‘moment”? In England decimalisa- 
tion of time may appear rather a remote topic, but it 
seems to have met with considerable favour at the Con- 
gress. The fact that a standing committee was appointed 
on the subject may not mean rapid progress, but M. 
Guyou mentions that his system has had a nine months’ 
trial on five French cruisers. 
Units form the subject of short papers by M. Lipp- 
mann, pp. 175-6, and Dr. Guillaume, pp. 179-183, and 
of a report by a special commission, pp. 184-6. M. Lipp- 
mann’s paper is theoretical, treating of various alterna- 
tives to the present second as unit of time. One is based 
on the Newtonian constant of gravity, a second is a 
submultiple of the sidereal year, a third is the time of 
vibration of a simple pendulum the length of which (at a 
given place presumably) would subtend a certain angle at 
the earth’s centre, a fourth is based on the oscillation 
period of acondenser. Dr. Guillaume’s paper is practical. 
He suggests the classification of watch movements 
according to diameter. Taking 2 cm. as point of 
departure, he suggests that the interval between suc- 
cessive classes should be 2 mm. above this point and 
1 mm. below. For balances he takes the formula 
a /1/M for a French (or half) vibration, where I is 
the moment of inertia, and M is the “ moment elastique ” 
(stiffness) of the spiral spring. He suggests that the 
number of the da/ance be the value of +I and the 
number of the sfxzzg be NM, both expressed in C.G.S. 
measure. These suggestions meet with considerable 
favour in the report of the special commission. The 
institution of definite types, with the elimination of 
intermediate sizes, is, of course, an important one for 
watchmakers. ( 
Amongst the papers bearing on topics of historical or 
current horological interest.may be mentioned those. by 
Rodanet on the proper definition of a chronometer, by 
Ditisheim on the classification of escapements, by Kaiser 
on the price and scientific value of chronometers, 
and by Caspari on the chronometers of. the’ French 
navy. In the paper by Ditisheim, pp. 40-46, there 
are a number of interesting data bearing on the 
NO. 1713, VOL. 66] 
comparative merits of different escapements. We have 
also a paper by A. Cornu, pp. 55-59, on the pheno- 
mena observed in magnetised watches, with a full dis- 
cussion of the effect of changing the position of a 
magnetised watch relative to the earth’s field; while 
Brillouin, pp. 164-174, treats experimentally of rapid 
variations in the amplitude of oscillation of balances, 
with special reference to the question of the shape and 
finish of the teeth of wheels. 
Amongst the papers dealing with instruments may be 
mentioned those by A. Cornu, pp. 47-54, on the clock at 
Nice, by Maillard Salin, pp. 63-5, on “‘ montres-a-billes,” 
by Féry, pp. 69-72, and Thury, pp. 146-152, on applica- 
tions of electricity, by Borrel, pp. 204-7, on a kind of 
Venetian blind semaphore for signalling time to ships, 
and by C. W. Schmidt, pp. 113-5, on his chronograph. 
This last instrument appears to be employed in France 
for measuring the velocity of projectiles, and is said to 
give velocities up to 700 metres a second correct to about 
I part in 500 A specially important paper is that by Dr. 
Guillaume, pp. 90-112, on nickel steels and their appli- 
cations to horology. The substance of this paper has 
mainly been published elsewhere, but it is presented here 
in a convenient form and it attracted considerable atten- 
tion at the Congress. Dr. Guillaume has yet another 
interesting communication, pp. 195-7, on an instrument 
for drawing the terminal curves of spiral springs in 
accordance with the results of Phillips’s well-known 
application of the mathematical theory of elasticity. 
The mathematical papers, though mentioned last, are 
by no means least in evidence. M. Faddegon, pp. 13-33, 
treats of the effects of changes of temperature on ordinary 
and on compensated pendulums. The formule he arrives 
at for the “grid-iron” pendulum are complicated and 
those for the mercury pendulum still more so.. In the 
latter case we encounter on p. 27 a determinant with ten 
rows and columns, and the mere look of the formulz on 
pp. 32 and 33 will probably suffice to damp the ardour of 
anyone anxious to combat the author’s conclusion, on 
p- 31, that it would be well for scientific purposes to 
give up attempts at compensation and revert to homo- 
geneous pendulums. 
M. Goedseels, pp. 73-89, treats of mathematical pro- 
cesses, less exhausting than least squares, for determining 
constants in linear formule: containing a considerable 
number of terms. Comparing the methods of Cauchy 
and of Tobie Mayer, he concludes that in point of sim- 
plicity the advantage rests sometimes with the one, some- 
times with the other, according to the circumstances of the 
problem. But in the case of the ordinary 6-term formula 
for the rate of chronometers he decides in favour of 
Mayer. 
The final mathematical memoir, pp. 217-252, consists 
apparently of a collection of already published papers by 
M. E. Caspari, the acting president, which the Congress 
decided to reprint. The common subject is the iso- 
chronism of helical springs. Calculations are made, after 
the methods of Phillips, Resal and Yvon Villarceau, of the 
influence of the “ centrifugal force” acting on the balance 
through its own motion, of the inertia of the spring, and of 
air resistance, friction, &c. There is also an investigation 
into the possibility of obtaining isochronism by varying 
