98 7 SCIENCE. 
winds in a cyclone, it may be an important 
aid to central warmth. 
Water-spouts are closely allied to tornadoes : 
but when seen in small form they approach the 
character of simple desert-whirls ; that is, they 
then depend merely on air warmed at the place 
where they occur, and not on the running to- 
gether of warm and cold winds from other re- 
gions. A probable cause for the excess of 
their strength above that of the sand-whirls lies 
in the smoothness of the water-surface on 
which they spring up, which will allow a long 
time of preparation; and in the moisture in 
the air, which will cause the warming of a 
greater thickness than if the air were very dry. 
_ The greater the thickness, the more their action 
will resemble that of a typical tornado. The 
appearance of the downward extension of the 
funnel-shaped cloud to meet the rising column 
of water is almost certainly only an appear- 
ance, and has the explanation already quoted 
from Franklin’s ingenious writings. 
We have relied largely, in the preceding ex- 
planations, on deductions from general prin- 
ciples, checked by the results of observation. 
The writings of many investigators have been 
examined, and in a few cases their names have 
been given; but the literature of the subject is 
now so extensive that full reference has been 
deemed unadvisable. Little attention has been 
paid to the older theories, in which conflicting 
winds and electricity were looked on as the 
chief causes of storms. ‘The latter is regarded 
as an effect rather than a cause; and, while the 
former has much importance when rightly con- 
sidered in connection with the earth’s rotation, 
it is of small value as originally stated, and is 
then limited to the production of short-lived 
storms in mountainous districts. The more 
important factors of the modern theory of 
storms are the consideration of the conditions 
of stable and unstable equilibrium of the at- 
mosphere, the true measure of the action of 
condensing water-vapor, the full estimation of 
the effect of the earth’s rotation, and the recog- 
nition of the necessary increase in the wind’s 
velocity as it is drawn in toward the storm- 
centre. W. M. Davis. 
THE CRITICAL STATE OF GASES. 
THE Philosophical magazine for August, 1883, con- 
tains a letter from Dr. William Ramsay which refers 
to observations upon the critical state of gases, pub- 
lished in the Proceedings of the London royal society, 
1879-80. The chief observations that had previously 
been made upon this interesting subject are those of 
Cagniard de la Tour (Annales de chimie, 2° série, 
‘ 
™ 
xxi. et xxii.), Faraday (Phil. trans., 1823 and 1845), — 
Thilorier (Annales de chimie, 2°™¢ série, 1x.), Nat- 
terer (Pogg. ann., xciv.), Andrews (Phil. trans., 1869). 
Andrews found that when a gas was compressed i in a 
closed space, and was maintained at a temperature 
below a certain limit, the pressure of the gas increased 
up to a fixed pointe beyond which condensation oc- 
curred. The pressure at which condensation takes 
place increases rapidly with the temperature of the 
gas. Atand beyond acertain temperature — the criti- 
cal temperature — no amount of pressure can produce 
any of the usual phenomena of condensation. The 
isothermal lines below the critical temperature are 
apparently discontinuous, one portion representing — 
no change of pressure corresponding to a change of 
volume. Above the critical temperature the isother- 
mals are continuous. 
The experiments of Dr. Ramsay were made upon 
benzine and ether, and a mixture of equal weights of 
benzine and ether. In one experiment a closed glass 
tube, somewhat in the shape of an hourglass, was 
used. One end of the tube was partly filled with 
ether, and was heated in an inclined position. The 
liquid expanded until, at the moment the meniscus 
disappeared, it nearly filled the lower half of the tube. 
On cooling, the liquid all condensed in the lower 
half, 
The experiment was varied by inverting the tube 
after the meniscus had disappeared. On cooling, 
the liquid condensed in the upper half of the tube. 
The tube was next maintained for some time at a _ 
temperature above that at which the meniscus disap- 
peared. On cooling, an equal quantity condensed in 
each division of the tube. It was observed, that, after 
the meniscus had disappeared, the part of the tube 
containing liquid had a different index of refraction 
from the other part. 
The conclusion to be drawn from these results is, 
that, at and above the critical point, the density of 
the liquid is the same as that of its saturated vapor: 
consequently, after a gufficient time, the liquid and 
its vapor will become mixed. Above the critical 
point, the surface tension of a liquid disappears. . 
This conclusion is confirmed by the experiments of 
M. Cailletet (Comptes rendus, Feb. 2, 1880). He found 
that when the lower part of his experimental tube was 
filled with liquid carbonic anhydride at a temperature 
of 5°.5, and the upper part was filled with air and gas- 
eous carbonic anhydride, a pressure of a hundred and 
fifty to two hundred atmospheres was necessary to 
cause the liquid to mix with the gas. At the sugges- 
tion of Mr. Jamin (Comptes rendus, May 21, 1883), 
hydrogen was substituted for the air in the upper 
part of the tube, and it was then found that a greater 
pressure was necessary to produce the mixture. This 
result would necessarily follow if we suppose that the 
mixture takes place when the densities of the liquid 
and the gas become equal. We cannot say that pr 
liquid is converted into gas by pressure. 
Though the densities of a liquid and its oteibtl 
vapor are equal, above the critical point, the two 
states of matter are still distinguished by other physi- a 
cal properties. Their indices of refraction are differ — 
[Vou. IIL, No, 51. 
