-2 June 2, 1923] 
NATURE 
739 

naturally concluded that they must be. One speaker 
confidently told me in the discussion that I was 
wrong, and that in the common toad the rugosities 
ave on the palmar surface! To show how the hands 
are placed I send a * peahcabery (Fig. 1) of a pair of 
Rana agilis killed and preserved while coupled. The 
lower digits of the male’s hands are the thumbs. 
Clearly the rugosities, to be effective, must be on 
the backs and radial sides of the digits, round the 
base of the thumb, as in our common frog, on the 
inner sides of the forearms, or in certain other posi- 
tions, but not on the palms of the hands. There are, 

Fic, 1. 
of course, minor variations, in correspondence with 
which the positions of the rugosities differ. The clasp 
of Alytes, for example, is first inguinal and afterwards 
round the base of the head (Boulenger). Minute thorns 
may be formed on the back of Bombinator and perhaps 
in other places on the skins of Batrachians, where they 
cannot serve as Brunftschwielen; but on the palm of 
Alytes they would be as unexpected as a growth of 
hair on the palm of a man. 
Dr. Kammerer’s own reply was on different lines 
from that of the speaker I have mentioned, but 
curious and, as I thought, significant. He asked us 
to note that in his lecture he had refrained from using 
the word “ Adaptation ’’—a defence sound perhaps, 
though surely disquieting to his disciples. 
The discoveries claimed by Dr. Kammerer are 
many and extensive. To geneticists that regarding 
heredity and segregation in Alytes (Verh. naturf. Ver. 
Briinn, r91r) which I called in question at the 
Linnean meeting is the most astounding. But what 
I then heard and saw strengthens me in the opinion 
expressed in 1913, that until his alleged observations 
of Brunftschwielen in Alytes have been clearly demon- 
strated and confirmed, we are absolved from basing 
broad conclusions on his testimony. 
May 16. 
NO. 2796, VOL. 111 | 
W. Bateson. 
The Light Elements and the Whole Number Rule. 
I HAVE recently developed a method of generating 
anode rays of high velocity which is much more 
suitable to mass-spectrum analysis than the hot 
anode method previously applied. By means of 
this device it is possible to obtain the mass-lines of 
the metals of the lithium and beryllium groups at 
the same time as those of such elements as carbon 
and chlorine, the masses of which are known. 
The masses of Li’, Li’, Be’, Na**, Mg™*, K®, K®, 
Ca*” have all been determined, and the divergence 
from whole numbers is in no case so great as one- 
tenth per cent. of the mass measured. The masses 
of the isotopes of lithium are most probably about 
9°005 of a unit high, but naturally this figure does not 
have much significance with the present apparatus. 
The effects with magnesium and calcium are too 
weak to show their fainter components, but the integral 
relations between these and the principal lines have 
already been demonstrated by Dempster. (Phys. 
Rev. xviii, xx.) 
This work completes the determinations of the 
masses of the more important isotopes of all the first 
twenty elements on the mass-spectrograph, and, 
with the obvious exception of hydrogen, each obeys 
the whole number rule to the accuracy of experiment, 
one part in a thousand. 
It is of particular interest that no difference in 
mass is detectable between the isobaric atoms Ca*® 
and A*, for general considerations might lead one to 
expect a radical difference in their nuclear structure 
owing to the presence of the two additional nuclear 
electrons in the latter. F. W. ASTON. 
Cavendish Laboratory, 
Cambridge, May 17. 

Microphonic Flames. 
[A FEw weeks ago it was reported in the daily Press 
that Dr. Lee de Forest had used a flame for the 
direct production of telephonic currents by sound 
waves. In response to a request for details of his 
device, Dr de Forest writes as follows.—Ep. Natrure.] 
», I have as yet prepared no paper on the subject 
of the ‘“‘ microphonic flame.’ For a long time I had 
puzzled over the problem of turning sound waves 
directly into electric telephonic currents. I recognised 
that sound waves passed through flames in the air ; 
also that a flame was, to a certain degree, conducting 
electrically. Hence, I reasoned that if one passed 
a current through a flame, its conductivity must vary, 
more or less, with the alternate waves of compression 
and rarefaction, which constitute sound. 
Setting out to verify my deductions, I succeeded 
almost at once. I employed first a “ bat-wing ”’ 
gas-flame, enriched this with potassium salts, used 
two platinum wire electrodes across a dry cell battery 
of 100 to 200 volts, in series with a high-resistance 
(radio) telephone receiver. By carefully adjusting 
the electrodes in the flame (especially the cathode— 
the position of the anode is not important; it can 
even be located a short distance outside the flame) 
I obtained in the telephone receiver a faint but 
very perfect reproduction of the music of a gramophone 
played 3 ft. from the flame. The adjustment of 
the gas pressure, using this type of flame, is critical. 
If too strong, the flame roars in the telephone 
receiver. If too low, the conductivity and sensitive- 
ness of the flame falls off. 
I next employed a type of Welsbach burner and 
mantle, using as electrodes platinum gauze “ im- 
bedded’ in the mantle and directly inside the 
mantle. Also, an oxy-acetylene flame, employing 
