i a 
May 2, 1912] 
NATURE 
207, 
extended source throws the shadow of a small object 
on a screen, under such conditions that the umbra of 
the shadow is not formed, then the shadow is the 
negative inverted image of the source of light. 
Another shadow phenomenon observed during the 
partial eclipse may be mentioned here, although its 
explanation is obvious. In cases where the leaves 
of trees were so far advanced that most of the sun’s 
rays were intercepted by them, the rays which passed 
through the small apertures between the leaves 
formed on the ground positive inverted images of the 
visible part of the sun’s disc. The oval patches of 
light seen on the ground beneath thick trees under 
ordinary conditions of sunlight are due to the same 
cause. During the eclipse, the rays of the sun re- 
flected from the free surface of water in a small glass 
formed a positive inverted image of the visible 
crescent of the sun on the walls or ceiling of the 
room in which the glass was placed. 
Epwin EDsErR. 
Halo during the Solar Eclipse of April 17. 
As Nature contains no mention of the circular halo 
that appeared for about half an hour during the solar 
interest, for possibly the appearance was very local. 
I was in the south of the Isle of Wight, at sea- 
level, and noticed, just as the air began perceptibly 
to cool, that a faint and very gauzy film of cloud 
collected round the sun. This was transparent 
enough for the sun to appear through it unmasked, 
but just dense enough somewhat to relieve the glare 
and make it possible to take hasty glances at the 
sun itself with the naked eye. After the clouds 
drifted into position, they remained through the 
whole time of the eclipse, but disappeared when the 
sun’s warmth returned. Shortly after their accumu- 
lation I saw a perfectly circular halo; this was 
coloured, but the bands of colour were only red, 
yellow, and greenish. The halo gradually increased 
in apparent diameter until it faded, as the eclipse 
waned. Marie C. Sropes. 
University College, London, April 27. 
The Smoke Problem. 
UnperR this head an anonymous notice appeared in 
Nature of April 11 of a little volume by Mr. Ruston 
and myself, in which the reviewer refers to certain 
““weaknesses in what is otherwise so excellent a 
work.” As the ‘‘ weaknesses” form the bulk of the 
review, I have permission to try to explain them. 
The first ‘‘weakness”’ is in reference to the origin 
of soot, which we ascribe partly to mechanical re- 
moval of dust and partly to incomplete combustion. 
The reviewer denies that coal-dust is a product of 
incomplete combustion, and also that tar and free 
carbon are formed in the destructive distillation of 
coal. I had imagined that tar and coke were among 
the principal by-products of the gas industry. 
The next criticism occurs in the paragraph that 
follows, and refers‘ to the amount of tar (we call 
“‘tar"’ the oil extracted by ether from soot and coal) 
in the original coals. The reviewer concludes, after 
citing some of our analyses, ‘‘surely the authors do 
not believe that a ton of these coals contains about 
a couple of gallons of ready-made tar.’ But suppose 
the authors have the weakness to accept the results 
of their analyses, what then? The reviewer offers 
no suggestion. 
In the next paragraph the reviewer finds fault with 
the statement: ‘‘The chimney gases were drawn off 
at the rate of about a litre a minute (i.e. through a 
NO. 2218, VoL. 89] 
narrow brass tube), which would approach the speed 
of the gases passing up the flue.’’ Although he is 
| good enough to interpret the passage for us in the 
| used that is the important factor.” 
only way in which it could possibly be interpreted, 
he concludes with the remarkable non sequitur that 
“if the flue draught was a litre a minute, it is no 
wonder their figures are abnormal.’’ Of course, 
there is no such suggestion that the flue draught was 
a litre a minute (which has no meaning, unless the 
area of the flue is known), nor is it so stated. 
Having made this gratuitous assumption, on what 
grounds does he conclude that our figures are 
abnormal? What are the normal figures? I believe 
that the figures of the late Sir W. Roberts-Austen 
and our own are the only records of the kind, and 
they substantially agree. 
In the next paragraph, among other remarkable 
items of information, is the statement that ‘‘the per- 
centage of soot to carbon burnt is of no practical 
importance. It is the percentage loss on the fuel 
The reviewer 
apparently fails to see that (1) the entire object of 
the experiments was to ascertain the amount of soot 
emitted, and (2) the percentage of soot on carbon 
| burnt can be easily calculated on the fuel used if the 
eclipse on April 17, the following facts may be of | 
amount of carbon in the fuel is known (as it was in 
every case). 
I do not wish to extend this reply by referring to 
our other ‘‘ wealxnesses,’’ which are of the same gross 
order. I can only thank the editor for his courtesy 
in giving me his permission to show how and where 
some of them, at least, may have had their origin. 
J. B. Couen. 
In the above remarks by Prof. J. B. Cohen on 
the review of ‘‘Smoke: a Study of Town Air,’”’ which 
appeared in Narure of April 11, the reviewer is first 
taken to task for denying that ‘‘coal-dust is a pro- 
duct of incomplete combustion, and also that tar and 
free carbon are formed in the destructive distillation 
of coal.’ The passage in the review was: ‘‘ Dust is 
not, as a rule, a product of incomplete combustion, 
nor is the tar and free carbon formed in the destruc- 
tive distillation of coal.’ The reviewer is still of 
opinion that coal-dust is not a product of incomplete 
combustion; by a strong chimney draught some coal- 
dust may be drawn up the flue, but it has certainly 
not been produced by combustion (unless Prof. Cohen 
looks upon the natural formation of coal as a process 
of incomplete combustion). Prof. Cohen elects to 
read the second part of the sentence as a denial that 
tar and coke are formed during gas manufacture, 
but it is doubtful if anyone else will do so; the 
reviewer’s statement is that the tar and free carbon 
formed in the destructive distillation of coal are not 
products of incomplete combustion. 
The authors give analyses of the original coal used 
in some of their experiments, and amongst the con- 
stituents of the coal figure certain percentages of tar, 
in one case amounting to 1-64 per cent.; and the 
reviewer says: ‘Surely the authors do not believe 
that a ton of these coals contains about a couple of 
gallons of ready-made tar.’’ To this Prof. Cohen 
replies: ‘“‘But suppose the authors have the weak- 
ness to accept the results of their analyses, what 
then? The reviewer offers no suggestion.” If the 
authors do believe it, I am afraid they would take 
any suggestion the reviewer could make as an 
impertinence. 
In replying to the criticism with regard to the rate 
of flow of the chimney gases in the flue, Prof. Cohen 
quotes from the book, and inserts five words which 
make the meaning clear, but which were not in the 
original paragraph. 
