Feb. 23, 1882] 
who wish to understand the whole subject from his volume, 
without reference to auxiliary text-books. To a slight 
extent he has revised the demonstrations and modified 
the diagrams of the Admiralty Manual—abridging these 
demonstrations so far as he thinks it possible to do so, 
He preserves Smith’s notation as gererally familiar. The 
newest part of the book is that in which he gives us the 
full details of the recent methods by which the amount of 
compensation can be determined with or without altitudes. 
In the account of these methods his readers will find col- 
lected much that it would be impossible for them to 
find in any English volume, and the exposition has all 
the merits of the French school. Smith and Evans 
showed us in their Manual how to determine the five 
essential coefficients upon which the deviation, when it is 
less than 20° in absolute value, depends at any position 
on the earth’s surface, and for any course of tke ship, 
three of these being constant. Two observations of the 
variation at any place accordingly suffice to determine the 
values of the deviation. Collet points out the necessity 
of compensation, that is to say, of the reduction of the five 
constants to insignificant values which can be determined 
even in a fog, when no observations of altitude can be 
taken, a method, of course, of the highest importance to 
practical men in circumstances of actual difficulty, which 
are constantly recurring. The theoretical part of the 
Admiralty Manual is given as succinctly as it is perhaps 
possible to give it if it is to be clearly mastered; the 
practical part is dwelt upon in full detail, and the rules 
are so simple and plain tbat ordinary captains in the 
merchant service ought to be able to use them accurately, 
even if they are unable to master the scientific part. 
There is an excellent account of Sir William Thomson’s 
compass. The fundamental ideas on which that instru- 
ment is constructed are that the magnetised needles must 
be so small that we may safely neglect their length, and that 
the intensity of their magnetism must be so slight that 
there is no reciprocal action between them and the soft 
iron correctors. The compass card is extremely light. 
A card of 1o inches diameter is directed by eight small 
needles, four on each side, like ordinary sewing-needles, 
of from 2 to 3 inches in length, and weighing in all about 
34 grammes. These are hung by two parallel silk threads, 
and attached to the card by silk threads passing through 
| the two eyes at the two ends of each needle. The entire 
| weight of the card, the needles, the outer circle of alu- 
| minium, the silk threads, the cap which rests on the 
| vertical point, &c., is only 12 grammes, which is about 
1-10th of what it is in the ordinary compass. This, of 
course, gives much less friction than usual between the 
point and the cap, so that the error due to friction is 
Weeduced in practice within a range of a quarter of a 
degree. 
The feeble magnetic moment of the system involves 
two important consequences—the period of oscillation 
round the position of equilibrium is only about 4o 
seconds, whereas it is three times as much in an ordinary 
compass card, and the suspension by silk threads makes 
| the whole card so elastic that it is much less liable to be 
| prejudicially affected by any sudden shock, such, for 
instance, as the firing of a cannon on board. 
| M. Collet gives an account of the compensated com- 
—<—————— — 
passes of Peichl, a lieutenant in the Austrian naval 
| 
NATURE 
383 
service. Peichl aims, like Sir William Thomson, to 
reduce the five coefficients used in determining the 
deviation to insignificant values. We must refer the 
reader to Mr. Collet’s book to show wherein Peichl differs 
from Sir William Thomson's system, and wherein M. 
Collet considers it to be markedly inferior to it in prac- 
tical value. 
The fifth part of the treatise discusses the methods by 
which it is possible, by the use of Sir William Thom- 
son’s deflector, to compensate and to obtain the neces- 
sary corrections for the compass in foggy weather, when 
no observations of altitude, celestial or terrestrial, are 
possible. These methods are admirable, and even on 
an iron vessel the most improved modern compass can 
be trusted by the mariner almost as completely as the 
chronometer its-lf. Just as in the case of the chrono- 
meter, however, it would be foolish to neglect the oppor- 
tunities of verification of the instrument which are 
constantly recurring on shipboard. Continually tested, 
and its performances brought frequently under review, it 
serves all the purposes of the seaman, and in foggy as 
well as in clear weather a captain can trust his compen- 
sated compass to navigate his ship. 
The brief historical exposition of the development of 
the compass, which occupies forty pages, is singularly 
interesting. Founded as it is on the Admiralty Manual, 
it is reasonable to expect to find, as we do, that the 
immense work of English men of science should be justly 
appreciated, but it is not perhaps so much a matter of 
course that that appreciation should be as generous as it 
is just. The French scientific man cannot always realise 
that science may sometimes emanate from other centres 
than Paris. M. Collet is not less scrupulously just to 
people of other nations than a German savant would he, 
he is more generous, and his book is more readable. 
OUR BOOK SHELF 
Stanford’ s London Atlas of Universal Geography. Quarto 
Edition. (London: Stanford, 1882.) 
Tuis new atlas appears to us to be superior in many 
respects to the ordinary run of such works. There are 
forty-four maps, and the selection has been made with 
great judiciousness, and with a special view to adapt the 
atlas to an English public. Britainand her dependencies 
occupy a prominent position ; Canada has three maps ; 
besides Australia there is a beautiful map of Tasmania, 
another of New Zealand, and one of the Fiji Islands, a 
specially original feature. Ceylon, moreover, has a map 
all to itself. The two maps devoted to Turkestan are of 
obvious utility, and have evidently been done with great 
care. There is a specially good separate map of Switzerland. 
Of Britain, besides the general maps of each of the three 
kingdoms, we have a fine orographical map showing by 
difference of tint both the varying height of the land 
and the varying depths of the sea around our shores ; and 
another map showing the distribution of the rainfall. 
There is a separate map of Japan, a very useful one of 
the Indian Archipelago, and a map of Africa in which 
several of the hitherto vaguely indicated Central States 
have had an approximate definition given to their areas. 
These are a few of the more prominent features of the 
atlas. The execution is on the whole thoroughly satis- 
factory ; several of the maps, indeed, were originally by 
Atrowsmith, Appended is a copious index of places, 
with their latitude and longitude. 
