3/2 
NATURE 
[Oct. 14, 1886 
Fired Gunpowder” (published in London, 1802, and recently 
reprinted, with his other essays, in America). 
Our Government official experiments give a pressure per square 
inch varying from 15°4 to 28°1 tons per square inch, the latter 
obtained in the 81-ton gun. Rumford’s maximum was 277 tons, 
as shown by overcoming the tested tenacity of metal that the 
powder tore asunder ; or 73 tons, as indicated by the lifting of a 
weight by the explosive energy. 
As I explained in the paper above named, the tearing test is 
fallacious as a theoretical measure of the force exerted, because 
the tenacity of the metal was tested by Rumford, as it still is by 
others, in most cases, by a gradually-applied strain, which should 
not be compared with a vibratory shock. As a measure of the 
practical bursting possibilities of gunpowder upon metal of given 
thickness and tenacity, as usually measured, Rumford’s figures 
are directly applicable, though allowance must be made for the 
relief afforded by the movement of the shot in a gun. 
My appeal was for a repetition of Rumford’s experiments by 
those who are responsible to the nation for these very serious 
matters, and for a reconsideration of the reliability of the method 
of testing by the ‘‘ Rodman” and ‘‘ Crusher” gauges, which 
have supplied such very different results from those of Rumford. 
My reasons for believing Rumford’s experiments to be more 
reliable than those of the Committee were stated as follows, and 
I still maintain their cogency :— 
“*(1) The resistance to be overcome, and by which the force 
was measured, viz. the gravitation of a known weight, was by 
far more definite and measurable than the degree of indentation 
or compression of a cylinder of copper, which serves as the 
measure of force in the Rodman and Crusher gauges. 
‘*(2) In Rumford’s arrangement the force of the explosion was 
more directly applied to the resistance by which it was measured 
than in the official experiments, where the shock of the explo- 
sion was first communicated to a solid piston 1 inch in length, 
and by this transferred to the copper cylinder of the Crusher 
gauge cr the knife of the Rodman gauge. By this arrangement 
much of the force is expended upon interne] work in the inter- 
vening piston, producing mechan‘cal vibration of its substance, 
and a returning wave of elastic compression, which would have 
no measurable effect on the gauge. Besides this, another por- 
tion of the force compressing the piston must be converted from 
mechanical motion into heat motion. 
“ Tf any reader supposes that I am hypercriticai in making this 
objection, let him try the following experiment. Take a block 
of iron—a common 1 Ib, weight, for example—place it on the 
hand, and the land upon a table ; then strike the weight smartly 
with a carpenter’s hammer. 
be struck upon the weight thus resting entirely upon the hand, 
and will scarcely be felt, provided the blows are dealt suddenly 
and smartly. The mountebank’s exploit of breaking a great 
stone upon a man’s bare breast, the common method of reducing 
the dimensions of geological specimens by holding them in the 
hand and cracking with a hammer, and the experiment of shoot- 
ing a bullet through a swinging door without moving it on its 
hinges, are familiar illustrations of this principle, which appears 
to have been overlooked in these official researches. 
‘* The complete absence of windage in Rumford’s arrangement, 
by exploding in a perfectly closed chamber, is a third advantage. 
I therefore regard Rumford’s experiments as the best that have 
yet been made on this interesting subject, although, as he himself 
admits, they are by no means free from error.” 
W. MATTIEU WILLIAMS 
Photographs of Stellar Spectra 
THE article upon this subject in NATURE, vol. xxxiv. p. 439, 
requires a correction which has been pointed out by Dr. Cope- 
land. The spectrum of the star DM. + 37° 3821 was observed 
by him on September 22, 1884, and found to contain bright 
lines ; the observation was published in the AZozthly Notices for 
December 1884, but was overlooked at the time when the article 
above mentioned was prepared. 
A similar correction, pointed out by Dr. Huggins, is required 
in the “‘ Investigation in Stellar Photography” by the present 
writer, published in vol. xi. of the ‘‘ Memoirs of the American 
Academy of Arts and Sciences.” On p. 208 the method of 
observing stellar spectra by means of a prism placed before the 
object-glass of a telescope is ascribed to Secchi. In fact, it had 
previously been employed by Fraunhofer. 
Epwarb C. PICKERING 
It will be found that blows which | 
would fearfully mutilate the hand if directly applied to it, may | 
The Late American Earthquake and its Limits 
IN your very flattering critique of my ‘‘ Alphabetical Cata- 
logue of European Earthquakes” the reviewer says :—‘‘ The 
tendency to alignment in volcanoes has often been noticed ; 
Prof. O'Reilly indicates a similar peculiarity in earthquakes, 
adding that the lines along which they range approximate to 
great circles. This inference or suspicion can be verified only 
by detailed charting.” Judging from the facts published up to 
the present relative to recent earthquakes of America and 
Europe, I think some such verification has been furnished by 
them. At the Exhibition of Scientific Apparatus held at South 
Kensington in 1877, I exhibited a globe mounted so as to allow 
of great circles being easily traced through points on the sur- 
face. Several coast-line great circles were shown thereon, 
amongst them that of the southern boundary of the Tertiary 
formation in the United States. It was also marked on the 
sketch earthquake-map of Europe exhibited before the Section 
of Geology of the British Association at their Swansea meeting 
of 1882, and on other maps, such as the earthquake map of the 
British Islands ; and yet no leading fact went to prove that any 
particular significance should be attached to this great circle. 
The earthquakes of August 27 and 28 in the United States have 
furnished, in my opinion, some proofs of this significance. The 
following are the places through which this great circle passes :— 
Victoria Fort, on coast of Gulf of Mexico ; Cairo (Ill.) ; axis of 
Lake Erie; Lake Ontario; River St. Lawrence (parallel to); New 
Brunswick coast of River St. Lawrence ; Labrador, south coast ; 
York Point and Straits of Belle Isle ; Ireland, Shannon mouth ; 
Wales, south coast of ; St. Bride’s Bay ; Mendip Hills ; South- 
ampton; Dieppe, north of; Chalons; Basle, north-east coast 
of Zurich Lake; Coire; Trent; Venice; Dalmatian coast ; 
south-west coast of Isola Longa; Mount Olympus; Skyro 
Island; Syrian coast, head of Akaba Gulf; Arabia, Mount 
Seiban, Wady Maifa; Cape Guardafui; Pacific Ocean, 
Paumota Group; coast of Mexico, near Cape Corrientes ; 
Zacatecas territory. 
According to Major Powell’s telegram, the origin of the 
earthquake was along a line of post-Quaternary dislocations 
on the eastern flank of the Appalachian Chain, especially 
where it crosses North Carolina. The great circle just de- 
scribed passes more inland than that mentioned by Major 
Powell, and was taken, as regards position, from the geological 
map of the United States, by C. H. Hitchcock and W. P. 
Blake, 1873, but it is parallel to the line limiting the Tertiary 
formation which crosses North Carolina, and which is probably 
also the seat of the post-Quaternary disturbance referred to. 
The great circle in question traverses the area of disturbance 
between Kennett (Ark ) and Buffalo (on Lake Erie). On the 
European side the following places lie near its direction. The 
Bristol coal-fields, where an explosion of fire-damp took place 
lately, about the time of the earthquake ; the English Channel, 
lat. N. 50° 10’ and long. W. 1° 49’, where an earthquake shock 
is reported to have occurred by H. Mohn in your issue of Sep- 
tember 23 (p. 496); the point lying about fifty-five miles to the 
south of the great circle, where it passes at Southampton. 
Switzerland: M. Forel reports in your journal of the 16th ult. 
(p. 469) a series of shocks in the western part of Switzerland 
having occurred in the first days of September, and which he 
considers as the szé¢e of the earthquake of August 27. In Eastern 
Europe an earthquake occurred on this same date, which tra- 
velled eastward from Malta to the South of Italy and reached 
Smyrna, which lies somewhat to the north of the great circle. 
In Mexico an earthquake is reported as having occurred at 
Tequisextlan on the 3rd ult. I can find no such place, but if it 
be the same as Tepantitlan, about fifty miles east-north-east of 
Guadalajara, it would be somewhat south of the great circle in 
question. As all these places are not far removed from the 
direction of the great circle, and as there must be several 
parallel lines of fissuring in the Appalachian Chain, thus forming 
a zone, there is in this way, I think, evidence furnished that a 
zone of seismic action exists, having the general direction of the 
great circle represented by the continuation of the boundary- 
line of the Tertiary formation in the United States to the west 
of the Mississippi Valley, as marked on the geological map of 
Messrs. Hitchcock and Blake. 
In the map forwarded herewith I have defined the surface of 
disturbance by lines joining the extreme points mentioned as 
having suffered shocks; but further information may, and 
probably will, modify this outline. The polygonal form thus 
obtained is, I think, more satisfactory than the curved forms 
