- 
JANUARY 17, 
1907 | 
NATURE 267 
&c., and it is precisely to matters like these that Prof. 
Lewes’s scheme of instruction is mainly directed. 
How to develop such a scheme on scientific lines 
under the conditions and limitations which have neces- 
sarily been imposed upon the authors is a problem of 
no small difficulty. That they have completely solved 
it in this book, even in the amended and extended 
form in which it now appears, they themselves would 
doubtless be the first to question. At the same time, 
there is nothing in our language even approximately 
resembling it, and it still remains the only manual 
which deals exclusively with the chemistry of the 
special matters with which the naval officer is more 
immediately concerned. 
The present volume differs from its predecessors 
many important points. With the collaboration of 
Mr. Brame it has been largely re-arranged and in 
great part re-written, and care has been taken to 
embody the latest information so far as this was 
available to the authors. 
The theoretical part is necessarily very restricted. 
Indeed, it is obvious that the philosophical aspects of 
chemistry have hardly more attraction for the authors 
than they have for the special class to whom the 
book is addressed. This, of course, is one of the 
difficulties of the problem. It is of no practical use 
to teach chemistry to naval officers as if they were 
going to be professional chemists, and the authors 
have therefore wisely confined themselves in the main 
to such theoretical aspects of the science as are re- 
lated to the matters with which naval men are directly 
interested. 
Naturally the applications of theory are constantly 
extending, and what is ‘‘theory’’ to-day becomes 
“ practice’? to-morrow—a truism of which the 
authors, it must be added, are not unmindful, for 
imperfect and partial as their treatment of doctrinal 
questions may be, it is probably sufficient for such 
developments as are likely, at all events in the near 
future, to occur. 
In one respect the book differs very materially from 
the ordinary run of chemical text-books, namely, in 
the large amount of original information it conveys. 
Prof. Lewes’s official connection with the Admiralty 
has necessarily caused him to pay special attention 
to chemical matters of importance to the service, and 
many of the results of his inquiries, some of which, 
indeed, are the outcome of prolonged investigation, 
are summarised in his manual. Although, as stated, 
it is primarily intended for the naval officer, there is 
much in the work which is of equal interest to the 
military man and to the practical engineer. 
We have noticed one or two errors, but they are of 
minor importance. Silicon was first isolated by 
Berzelius in 1523, and not, as stated, by Davy in 
1813. There is no such thing as P,O,; phosphorous 
oxide has the formula P,O,, just as its analogue 
arsenious oxide has the formula As,O,. The descrip- 
tion of the mode of manufacture of the lucifer match 
hardly corresponds with modern practice; white phos- 
phorus is rarely used in the “ strike-anywhere ’’ match 
to-day; it has been almost wholly replaced by phos- 
phorus sesquisulphide. 
NO. 1942, VOL. 75] 
in 
| 
| 
| 
| 
| 
FORMATION OF ICE ON THE 
ST. LAWRENCE. 
Ice Formation, with Special Reference to Anchor- 
ice and Frazil. By Howard T. Barnes. Pp. x+ 
260. John Wiley and Sons; London : 
(New York : 
Chapman and Hall, Ltd.) Price 12s. 6d. net. 
HE effects of the severe Canadian winters on 
such a large river, with a variable flow, as the 
St. Lawrence afford remarkable opportunities for 
studying the phenomena of ice formation, which the 
author has availed himself of during the last ten 
years; and this book records the results of his observ- 
ations and the conclusions he has drawn from them. 
Moreover, in order to render the account of his in- 
vestigations on ice formation more complete, he has 
introduced the subject with three chapters, on the 
“Physical Laws governing the Transfer of Heat,” 
‘Physical Constants of Ice,’’ and ‘‘ Formation and 
Structure of Ice,’? and has added towards the end a 
chapter on previous ‘‘ Theories to account for Frazil 
and Anchor-ice,’’ which would more suitably have 
preceded the exposition of his own views in the fourth 
chapter. Lastly, in the final chapter, the author gives 
a practical application to his investigations by in- 
dicating the causes which, in severe winters, are liable 
to occasion the stoppage of water-power works, and 
suggesting measures by which accumulations of ice 
at critical points, tending to arrest the flow of water, 
may be reduced. 
Three forms of ice are found in the St. Lawrence 
in winter, namely, sheet- or surface-ice, frazil-ice, and 
anchor-ice, differing in their mode of formation, their 
general appearance, the position they occupy in the 
river, and the effects which they produce. Sheet-ice 
is the well-known form of ice which, when the 
temperature falls below the freezing point, gradually 
forms on the surface at the sides of a sheet of still 
water, spreads out into deeper water if the cold 
continues, and increases slowly in thickness. Frazil- 
ice is the French-Canadian term, signifying cinder- 
ice, for a peculiar spiky form of ice supposed to 
resemble cinders, which is formed on the surface in 
open channels where the current is too rapid for the 
border-ice to extend across them; and this ice, which 
is sometimes called slush-ice, varies in its formation 
according to the amount of agitation of the water, 
appearing as flat plates on the smooth surface of a 
current, or as numberless minute needles in rapids 
and at the base of waterfalls. These needles of ice 
increase in bulls in traversing open water for some 
distance, and eventually fill up the open channel 
during the prevalence of great cold and strong winds 
often experienced in Canadian winters. This fine ice 
is often carried by the current a long way under 
surface-ice which has formed lower down, and, be- 
coming attached to the under-side of this sheet, and 
gradually accumulating and becoming consolidated, 
it is liable to dam up the channel completely down 
to observed depths of 80 feet, and, consequently, raise 
the level of the river considerably above. 
Anchor-ice, as it is called in Canada, and known 
as glace-du-fond in France and Grundeis in Germany, 
has been long observed in most countries where ice 
