386 
NATURE 
[JANUARY 18, 1917 
general properties of errors are treated, the dis- 
tinction between errors and mistakes is carefully 
explained, and the reader will derive from these 
few pages a very clear idea of the kind of errors 
with which it is the mission of least squares to 
deal. This is followed by a chapter on the general 
theory of probability treated algebraically, and 
containing a brief explanation of the application 
of the theory to practical problems, such as life 
insurance, together with some examples from 
mortality tables. : 
We then reach the most important part of the 
book, chaps. iv. and v., comprising an exposi- 
tion of the theory of errors and least squares on 
very elementary, but quite orthodox, lines. The 
great variety in the problems introduced to illus- 
trate the text is very noteworthy: statistical 
tables, electrical resistance, balance constants, 
volumetric solutions, specific gravity bottles, 
surveying, transits of stars, the resolution of 
apparent parallax into actual parallax and proper 
motion, the solubility of salts, are all made to 
serve. The problems on chemical work are par- 
ticularly suggestive, while the one on locating a 
distant station in surveying is of interest in 
employing rectangular co-ordinates instead of 
angles. There is an unfortunate mistake in the 
first numerical example of a normal equation 
(p. 75); the right-hand side of the equation should 
read 3676, and the results of the problem as 
given in the text are appreciably inaccurate. 
-Chap. vi. is on empirical formule, and 
includes some useful hints as to the choice of 
mathematical expressions to represent the un- 
known relations between variables. A problem 
on the “reduction of pendulum to zero arc,’’ on 
pp..107 to 110, in which time is measured to the 
millionth of a second and arcs are recorded in 
whole degrees only, looks rather uncanny, but 
may be unexceptionable. The next chapter is on 
weighted observations, and follows the usual lines. 
In the final chapter, on the general theory of pre- 
cision, an elementary knowledge of integration is 
assumed. The appendix, to which, as already 
mentioned, the more difficult analysis is relegated, 
contains also a very complete table of formule, 
all of which have been deduced in the text. On 
the whole, this is a good book, and being far less 
mathematical than most other works on the sub- 
ject, it is likely to appeal to a wider class of 
readers. 
NOXIOUS INSECTS. 
Medical and Veterinary Entomology. By W. B. 
Herms. Pp. xii+393. (New York: The Mac- 
millan Company; London: Macmillan and Co., 
Ltd., 1915.) Price 17s. net. 
1 Wait excellent text-book is based on manu- 
script used in teaching in the University of 
California and in the San Francisco Veterinary 
College. It is not intended to be a very com- 
prehensive treatise, but an attempt to systematise 
the subject. It, however, goes beyond this, as 
new matter is here and there incorporated, thus 
making the volume of greater value. It is mainly 
NO. 2464, VOL. 98] 
adapted to the American continent, but will be 
found of general usefulness elsewhere. For 
instance, there is only a key of the North American 
genera of ‘Tabanide. ‘The first chapter is an 
introduction. The second deals with parasites and 
parasitism in general; the third with insect 
anatomy and classification, with a useful working 
key to the orders of insects. The mouth-parts are 
shortly but very concisely treated in chap. iv. ;. 
this portion might well have been amplified. 
Cockroaches, beetles, and thrips are dealt with, 
and the small yet important part played by cock- 
chafers in the spread of Echinorhynchus gigas and. 
the uses of Spanish fly, etc., are concisely detailed. 
There is an interesting chapter on lice (pp. 52-68) ; 
we notice here that the human clothes louse is still 
called Pediculus vestimenti instead of P. humanus ; 
the figures given here are not good. Bed-bugs 
and cone-nose bugs form the theme of chap. viii. 
An excellent précis on mosquitoes or Culicidz is 
found in chap ix. (pp. 89-100), the classification 
used being that of Theobald and others, and not of 
the American dipterologists. | Mosquito-carried 
diseases and contro] are also explained, and a full 
key of classification given under the Theobaldian 
system. 
Other blood-sucking flies are dealt with, such as 
the buffalo-gnats, or Simulidz, and horse-flies, or 
Tabanidz, and notes on their control and relation 
to diseases are given. Naturally, the house-fly is 
fully described, twenty pages being devoted to its 
life-history, habits, and its relation to diseases, and 
another twenty-two pages to its control. The 
African tsetse-flies, or Glossinz, and the horn- and 
stable-flies, are also fairly fully dealt with in 
chap. xv. (pp. 207—232). 
An interesting account of Myiasis is given, in- 
cluding attacks of flesh-flies and bot-flies, or 
Gistridee, and others, such as the Congo floor- 
maggot and the West Indian and American screw- 
worm. The portion dealing with the ox warble- 
flies, pp. 251-254, is not quite up to date; for 
instance, it is said that the larve are licked off 
by the tongue, and so pass into the cesophagus, 
Carpenter’s researches in Ireland evidently being 
unknown to the author; 
that the larve enter by the skin, especially 
of the legs, and it is unlikely that any enter 
as described in this work. Nothing is sai 
of. their attack on human beings, the so-called 
“creeping disease,” which is frequent in some 
countries. The remainder of the work is taken up 
with chapters on fleas and louse-flies, ticks, mites, 
including scab in sheep, scaly lege in fowls, and 
itch, and also an account of venomous insects and 
arachnoids. The section on louse-flies (Pupipara) 
might well have been extended; the account of the 
sheep “ked ” is very brief,. whilst all that is said of 
the Hippoboscide is contained in five lines, deal- 
ing with H. equina. Fuller information on Pedicu- 
loides ventricosus might also have been given, and 
the recent work of Willcocks in Egypt and others. 
included. 
The work ends with a four-page appendix deal-. 
ing with general classification of bacteria and 
protozoa. Poe 
these clearly prove — 
oe 
