194 
valves of different descriptions we find much valuable 
information, including the piston type of valve, largely 
used at sea, and now, under the name of Smith’s Patent, 
being experimented with on the North-Eastern and Mid- 
land Railways, the Highland Railway having tried and 
discarded it some short time ago. Valve gear in detail 
comes next, and we cannot agree with the author when 
he says that in inside cylindered engines expansion links 
with one bracket only are used. This is the exception 
and not the rule—v¢de the practice of the Brighton, 
Caledonian, and North British Railways. 
Taken as a whole, the descriptions of all detailed work 
represent modern practice, and the engineering student 
will find much to learn in these pages. Chapter xiii. 
deals with the all-important question of the general con- 
struction and design of the locomotive boiler; after 
discussing the questions which really concern its dimen- 
sions, the question of various types is described. The 
now fashionable Belpaire type is badly illustrated in 
Fig. 155, which must represent an American or con- 
tinental type of boiler, although the author does not 
say so, and the arrangement of stays and plating is 
certainly not of British design. The Belpaire boilers 
designed by Messrs. Neilson, Reid, and Company for 
the Mexican railways some years ago might be taken 
as fair representatives of this type of British design, and 
should be illustrated in a future edition—they being the 
prototype of some running on more than; one British 
railway. 
In order to allow freedom for expansion of the tube- 
plates of Belpaire boilers of British design, it is usual to 
arrange the last transverse rows of vertical stays so that 
any vertical movement of the fire-box will not be trans- 
mitted to the wrapper-plate ; moreover, a similar arrange- 
ment might be placed above the door-plate with 
advantage. 
On the use of steel in boiler construction we find much 
valuable information, but we most distinctly disagree with 
the author when he states on p. 200 that steel of boiler- 
plate quality contains a maximum of 15 per cent. anda 
minimum of to per cent. of carbon ! What has happened 
to the decimal point? The author, like many others, has 
not yet got over the idea that steel-plates require very 
special treatment in the flanging-shed and boiler-shop. 
This is not the case; steel-plates, as manufactured to-day, 
are more uniform in quality, and are certainly as easily 
worked as Yorkshire iron. On p. 215 we read that on 
the Caledonian Railway the roof of the fire-box is sup- 
ported by vertical stays fastened in series of threes ina 
longitudinal direction. These stays were on the scrap- 
heap years ago. 
On the question of machine riveting, our author 
maintains that subsequent caulking of rivet-heads is 
unnecessary. If this is the case, why do our best firms 
of locomotive builders invariably carefully caulk every 
rivet-head before the boiler is tested? They work for a 
profit, not for honour and glory. 
Mr. Ravenshear gives a very full description of con- 
tinental and American locomotive practice, with illus- 
trations, which will be found towards the end of this 
volume, besides the usual descriptive accounts of the 
vacuum and Westinghouse railway brakes. Taken as 
a whole, this work is one of the best of its kind 
NO. 1548, VOL. 60] 
NATURE 
[JUNE 29, 1899 
that has been published on the subject. The strains 
experienced by various parts of a locomotive during 
work are impossible to calculate, and, therefore, it must 
be every-day experience that can train the successful 
designer. For this reason a text-book on this subject 
can only be descriptive of work done which successfully 
withstands the usages of every-day work. 
NORMAN J. LOCKYER. 
THE HEREFORD EARTHQUAKE OF 1806. 
The Hereford Earthquake of December 17, 1896. By 
Charles Davison, Sc.D., F.G.S. Pp. xi + 303. 
(Birmingham: Cornish, 1899.) 
PALER an interval of more than two years Dr. 
Charles Davison has at last given us, in a volume 
of 303 pages, his long-promised account of the earth- 
quake which, in the early morning of December 17, 
1896, rudely awakened the inhabitants of the Severn 
Valley. 
When we look at the 2902 epitomised accounts which 
Dr. Davison has brought together respecting an earth- 
quake which in many countries would have been regarded 
with as much indifference as a sprinkling of rain, we are 
inclined to ask whether the examination of this long 
series of remembrances, obtained from a community 
more or less excited by phenomena with which they had 
but little experience, would be likely to lead to results of 
any value. Had this earthquake originated in a sparsely 
populated country where there were difficulties in obtain- 
ing accurate time, the analysis of observations taken 
under such conditions would, to a large extent, have been 
labour in vain. . 
Although no special provisions are taken in Britain 
for the observation of earth tremors, as the one now 
under consideration occurred at the waking hour of 
many millions of people who, lying on their beds, were 
in the best possible position for noticing slight vibrations, 
and for the most part had the means of obtaining fairly 
good time, and above all were intensely interested in 
the phenomenon they experienced, the conditions for 
obtaining a large series of valuable records were un- 
usually favourable. Within the epifocal area where 
chimneys fell or were “hurled to some distance ”—which 
we doubt—and buildings were unroofed, and within at 
least one hundred miles of the same, all the observers 
had but little doubts as to the nature of the movements 
they experienced. Beyond these limits in very many 
instances it is likely that many observers only realised 
and remembered that a something or other had rattled, 
perhaps the window or a lamp-shade, after they had read 
their morning papers, and with feelings of satisfaction 
as participants in an alarming disaster, they threw in 
their notes and helped to complete an important chapter 
in British seismology. If every time a window was 
slightly shaken, glasses rattled, or other unaccountable 
microphonic disturbances were perceptible could be re- 
corded, and the collected results analysed, it is extremely 
probable that the seismic register for the British Islands 
would be considerably increased. 
Before discussing the catalogue of observations, Dr. 
Davison sets out by showing that there is a reality of 
connection between the majority of earthquakes and the 
