Oct. 29, 1885] 
NATURE 
625 
easterly current over the N.W. monsoon and of an upper current 
over the S.E. trade, more southerly than the surface wind, is 
not only altogether new, but also quite anomalous. 
In Australia, and the Southern Hemisphere generally, the 
upper current over a N.W. wind is from about W. and over a 
S.E. wind from about E. 
On my way home I ran a section across the Atlantic from Rio 
to Teneriffe, but the absence of cirrus prevented any satisfac- 
tory determination of the upper winds in that region. 
The matter is, however, so important that I start again ina 
few days for the hurricane region of Mauritius, where I hope to 
observe one of these exceptional cyclones. Then I hope to 
repeat a section of the Indian Ocean between Mauritius and 
Bombay, and afterwards, if all goes well, to get some sections in 
the Pacific to see what the meaning of this curious discovery 
may be. RALPH ABERCROMBY 
21, Chapel Street, S.W., October 26 
The Hellgate Explosion and Rackarock 
THE statement in NATuRE of the 15th inst. (p. 575) that 
rackarock is ‘‘blasting gelatine” or ‘‘ nitroglycerine with 
compressed gun-cotton” is incorrect. Rackarock is simply 
powdered potassium chlorate, impregnated with an inexpensive 
oily combustible, such as coal-tar oil, and is one of my safety- 
explosives, which I discovered in 1870, patented in England, 
April 6 and October 5, 1871, and described more fully in the 
Fournal of the Chemical Society for August, 1873, under the 
title : ‘On a New Class of Explosives, which are non-Explosive 
during their Manufacture, Storage, and Transport.” 
I am not responsible for the quaint name which the Americans 
have been pleased to give to my child. 
As the so-called ‘‘ rackarock” is not very sensitive or easy to 
explode, it requires a strong primer or detonator to set it off. 
This property, which I have fully discussed and particularly 
accentuated in my paper of 1873, explains why Gen. Newton, 
the Chief Engineer of the Hellgate mine, took the precaution 
of placing as a primer such a powerful charge (33 tons) of ex- 
pensive dynamite on the cheaper charge of the potassium 
chlorate mixture (107 tons), a precaution carried here perhaps a 
little too far. 
Still it is satisfactory to see that my safety-explosive performed 
the main part of the labour and rendered good service in the 
advancement of the works of peace. H. SPRENGEL 
Savile Club, 107, Piccadilly 
[We are very pleased to insert Dr. Sprengel’s correction as to 
the composition of ‘‘rackarock.” Up to the time of our notice 
about the explosion going to press the only information we could 
obtain was that it was the same substance as blasting gelatine, 
but with a less portentous name.—ED. ] 
An Earthquake Invention 
In your number for October 15 (p. 573) your numerous scien- 
tific readers will be interested to find a pretty long letter under 
the above heading from so able a seismologist as Prof. John 
Milne, of Tokio, Japan. Yet, his invitation notwithstanding, I 
must decline any discussion with 42m, either about my old letters 
which he refers to, or his own much changed opinion on their 
subject, since the occasion for my writing them occurred. 
Those points, Mr. D. A. Stevenson, who is also invited, may, 
or may not, take up. My letters were impersonal, and dealt 
only with a British Association Report. I desire also to con- 
tinue to keep them strictly to that, even to the very words of the 
particular Report as given forth to the world with all the usually 
unquestioned authority of that mighty Association, in their B.A. 
volume for 1884, p. 248, Section entitled ‘‘ Experiments on a 
Building to resist Earthquake Motion.” 
C. PIAzzI SMYTH 
15, Royal Terrace, Edinburgh, October 16 
On the Behaviour of Stretched Indiarubber when 
Heated 
SoME time ago (NATURE, vol. xxv. p. 507) you permitted me 
to express a doubt as to the invariable success of an often-quoted 
experiment with cylinders of bismuth and iron, intended to 
Illustrate some relations between specific heat and thermal 
conductivity.1 I regret that I have made further progress along 
1 Perhaps I may mention in passing that if lead is substituted for the 
bismuth the experiment succeeds perfectly, as theoretically it should do. 
the evil road of scepticism. I should like, in fact, to ask 
whether it is absolutely true to say without qualification, as is 
done in many text-books, that india-rubber (when stretched) 
forms an exception to the general law that the volume of a body 
is increased when the temperature is increased. The usual 
form of the experiment supposed to prove this is well known: a 
piece of india-rubber tube or cord is stretched by a weight con- 
nected with a long light index-lever, and it is shown that when 
heat is applied the india-rubber gets decidedly shorter. 
I have always had some hesitation in showing and explaining 
the result of the experiment in the above way, especially as I 
could not find any proof given that the contraction in length was 
not compensated, or more than compensated, by an expansion 
in other directions (like that of a worm in its creeping progress, 
or of a dry rope when wetted). I had, in fact, lately arranged 
an apparatus for determining the coefficient of expansion of india- 
rubber, whether positive or negative, when I found that the sub- 
ject has been very fully investigated by Dr. J. Russner, of Chemnitz 
(see Carl’s Repertorium for 1882, pp. 161 and 196). 
His results are briefly these :— 
(1) That india-rubber (of which several kinds were examined) 
has without exception a definite coefficient of expansion which 
is always positive ; experiments made at temperatures varying 
from 0° to 53°°4 gave, for its value at 10°, 0°000657; at 30°, 
0000670. 
(2) That india-rubber in a stretched state expands to the same 
extent as when it is not stretched. No point of minimum 
density was observed, such as Puschl supposed to exist. 
(3) That the apparently anomalous behaviour of stretched 
india-rubber when heated is simply a case analogous to those of 
anisotropic crystals, which expand to different extents in different 
directions. Iceland spar, for instance, as Mitscherlich showed, 
actually contracts in a direction at right angles to its principal 
axis when heated, although its volume is, on the whole, 
increased. 
Although ordinary india-rubber is, of course, isotropic, yet 
when stretched it becomes anisotropic, as may easily be shown 
by stretching a piece until it is semi-transparent, and placing it 
between crossed Nicols ; the direction of the strain lying at an 
angle of 45° with the plane of polarisation. Distinct colours, as 
with a selenite film, will be seen, varying from red to blue with 
the amount of strain. 
The fact that india-rubber becomes hot when stretched, and 
especially if stretched and allowed to contract several times in 
succession, may perhaps be accounted for by molecular friction. 
It would almost seem, then, that in the account given in many 
books the truth, as well as the india-rubber, has been slightly 
“stretched.” H. G. MADAN 
Eton College, October 23 
The Resting Position of Oysters 
IN carrying out a series of experiments on the artificial breed- 
ing of oysters in my private aquaria, I noticed that the young 
oysters born in the tanks rested on the flatter shell when they 
obtained a flat surface, such as a tile, to adhere to, but when I 
so arranged that they had irregular surfaces to deal with, 
such as little bundles of twigs, some adhered one way, and some 
the other. But where young oysters, nearly two years old, were 
moyed from their original supports, and were compelled to find 
new ones, they selected the flat shell to rest upon in every 
instance, except where they were placed on sand, in which case 
they rested on the convex shell, in order apparently to avoid 
clogging the mouth of the shell with sand. Is it not possible 
from these observations that adult oysters vary their position 
according to the nature of the ground they are on. I have 
seen adult oysters on muddy ground lying on the convex shell, 
while where adhesion to a flat surface could be obtained, they 
were all on the flat shell, and pectens are dredged with Balari 
and other growths on the flat shell in some instances, and on 
the convex shell in others, principally, however, on the latter. 
H. STUART-WORTLEY 
South Kensington Museum, October 23 
The Value of the Testimony to the Aurora-Sound 
I HAVE read with much interest the descriptions of this sound 
as given by Dr. Sophus Tromholt’s correspondents in NATURE of 
September 24. I was, however, struck by the similarity. of 
these descriptions to the well-known phenomena of tinnitus 
