APRIL: 18, 1907 | 
NATURE 
583 
rather as unhealthy accretions. At the same time, 
the author, following Gauss, has failed to recognise 
the extreme diffidence with which Laplace put  for- 
ward his hypothesis. By many, the caution and re- 
serve with which Laplace accompanied his sugges- 
tions will always be regarded as a model of good 
taste and evidence of a correct scientific attitude. 
The New Hygiene. By Elie Metchnikoff. Pp. viii+ 
104. (London: William Heinemann, 1906.) Price 
2s. 6d. 
Tuts little book contains the three Harben lectures 
delivered by Dr. Mefchniloff at the Royal Institute 
of Public Health last year, an appreciative preface 
being contributed by Prof. Ray Lankester. The 
‘“ Hygiene of the Tissues’? is the title of the first 
lecture, and in it the phenomenon of phagocytosis is 
discussed at some length, and since this fact is con- 
sidered to be the principal means of defence of the 
body against the invasion of microorganisms, and 
since such drugs as alcohol, opium, and many others 
impede phagocytosis, it is concluded that their use 
should be avoided or limited in the treatment of 
disease, and certain substances such as blood serum 
and salt solution. which stimulate phagocytosis, 
employed in certain circumstances. In the second 
lecture, on the hygiene of the alimentary canal, 
the evil effects of parasitic organisms are dealt with, 
and the use is advocated of sterile food so far as is 
possible. The third lecture deals with hygienic 
measures against syphilis, and the use of inunction 
of mercurial ointment as a prophylactic against in- 
fection detailed. The book is of extreme interest, and 
one that should be widely read by the educated public. 
IGE AER al 
Synopsis of Mineral Characters. Alphabetically 
arranged for Laboratory and Field Use. By Ralph 
W. Richards. Pp. v+97. (New York: John Wiley 
and Sons; London: Chapman and Hall, Ltd., 1907.) 
Price 5s. 6d. net. ; 
Tue title of this convenient pocket-book serves to 
define its scope. Emphasis is laid upon crystal form, 
habit, system, cleavage, hardness, fusion, and solu- 
bility in hydrochloric or other acid. Definitions of 
mineral terms and of rocks associated with the 
minerals included are also provided. The arrange- 
ment of the matter makes reference to the book easy. 
LELTERS TO THE EDITOR. 
[The Editor does not hold himself responsible for opinions 
expressed by his correspondests. Neither can he undertake 
to return, or to correspond with the writers of, rejected 
manuscripts intended for this or any other part of NATURE. 
No notice is taken of anonymous commiunications.] 
Anomalous Dispersion and Ionisation 
My criticism of Mr. Schott’s interesting experiment 
(Nature, March 14, p. 461) was due to my carelessness in 
reading his brief account. I failed to notice the words 
““and the tube ’’ in his description of the battery connec- 
tions, which fact, together with the low voltages which 
he used, gave me the impression that he employed the 
current to heat the wire, the ions being derived from the 
hot metal. 
The experiment as actually performed is of considerable 
interest. Everything depends upon just what is meant by 
““the dispersion is completely annulled.’’ I infer that the 
oppositely curved branches of the spectrum move back 
into the original straight line, but nothing is said as to 
whether the gap caused by absorption is filled in. The 
dispersing power of the sodium tube depends upon the 
density gradient of the sodium vapour, as we pass from 
the floor to the roof of the tube. Anything which inter- 
feres with this will alter the dispersion. If the discharge 
stirs up the vapour and renders it homogeneous over each 
WO 1955, VOL. = 5 
| Physical 
cross-section of the tube, the dispersion will be annulled,. 
while the absorption will remain. The nature of the 
action going on in the tube can probably be learned by 
looking through the tube at a bright sodium flame, or a 
gas flame ‘provided with a blue screen which transmits 
only the region 4600-4900. It will be found that the upper 
portion of the tube is fairly transparent to the radiations, 
while the vapour along the floor is quite opaque to them. 
The effect of the current on the transparency at different 
levels should be noted. 
There has always appeared to me to be some mystery: 
about the behaviour of sodium vapour in highly exhausted 
tubes, for it is difficult to see how equilibrium can exist 
between the dense vapour along the floor and the nearly 
perfect vacuum along the roof. In the light of recent ex- 
periments which I have been making, I now believe that 
I have found the solution of the apparent difficulty. The 
actual density of the vapour along the floor has in all 
probability been over-estimated. It is usual to exhaust 
the tubes to a pressure of a millimetre or two. In all 
probability, pure sodium vapour at two millimetres pressure 
is what we should call a very dense vapour (considered! 
optically). Suppose, now, we heat the floor of the tube 
to the temperature at which the vapour pressure of sodium 
is equal to the pressure of the residual gas in the tube. 
The density of the sodium vapour considered alone (partial 
pressure) will depend upon the rate at which it can diffuse 
through the residual gas to the cooler roof of the tube. 
If the sodium vapour is given off from the molten metal 
more rapidly than it can diffuse away, we may have pure 
sodium vapour at the surface of the metal, and mixtures 
of sodium and hydrogen in decreasing proportion as we 
pass upwards towards the roof, the total pressure being 
the same at every point, however. If this is the true state 
of things, the dispersing power of the tube would dis- 
appear if every trace of the residual gas was removed. I 
intend shortly to test this point. I have already found 
that in the long steel tubes such as are used in observing 
the magnetic rotation of the vapour, the density of the 
sodium vapour is greatly increased by the admission of 
hydrogen or air. In this case the central portion of the 
tube is uniformly heated with an electric oven, and the 
sodium distils away to the cooler portions. The presence 
of hydrogen or nitrogen hinders this process, the gas hold- 
ing back the sodium vapour, so to speak, and allowing it 
to acquire a density, or rather pressure, equal to its own. 
This way of looking at the state of affairs in the tube 
may prove helpful. in explaining the interesting effects 
observed by Mr. Schott, whose further experiments I shall 
follow with interest. R. W. Woop. 
Baltimore, April 2. 
Positive Streams in ‘‘ Crookes”’ Tubes. 
REFERRING to the abstract of Mr. F. W. Aston’s very 
interesting paper read before the Royal Society on 
December 13, 1906, ‘‘ On Experiments on the Length of 
the Kathode Dark Space with Varying Current Densities 
and Pressure in different Gases,’’ published in your issue 
for April 11 (p. 574), may I point out in reference to the 
therein contained statement that ‘‘ the stream of positive 
ions may be strikingly shown by a rotatory mica mill 
mounted inside the dark space which rotates violently in 
the opposite direction to the familiar ones designed to 
show the motion of kathode rays away from the electrode,”’ 
that in my two papers “‘ On the Circulation of the Resi- 
dual Gaseous Matter in a Crookes’ Tube,”’ read before the 
Society, and published in the Philosophical 
Magazine for October, 1898, I showed similar results, 450: 
that mica mill wheels which turned in one direction under 
kathode-ray bombardment, turned in the opposite direc- 
tion when so placed as to be iust outside of the stream 
of kathode rays, thus indicating a current of particles 
proceeding towards the kathode, which particles I found 
to be charged positively ? 
These results, as stated in my first paper, could only be 
obtained with extremely high vacua, when no doubt the 
mica mill wheels were inside the dark space, as is found 
necessary by Mr. Aston. 
A. A. CaMPBELL SWINTON. 
66 Victoria Street, Westminster, S.W., April 13. 
