Marcu 21, 1907 | 
MAT OF 
487 
blamed, 
reasons, 
especially because he had omitted, for 
to add carbolic acid to the prophylactic. 
alarm was produced. The idea that the poisoning 
due, not to local accident, but to carelessness at the 
atory, caused, I have been told, a sudden and wholesale 
rejection of the invaluable vaccine by the people, with 
the probable result that thousands of lives may have been 
lost from plague. 
Now it appears that the tetanus bacillus could not have 
entered the bottle at the laboratory at all! I agree with 
Prof. Simpson (British Medical Journal, February 9) in 
thinking that the arguments on this point are extremely 
strong. Had the contents of the bottle been polluted at 
the outset, they would have had a very offensive smell 
when used some time later, and would have produced a 
very rapid infection in the inoculated. As a matter of 
fact they had no smell, and produced a slow infection, 
while bottles filled simultaneously were quite sound. 
Moreover, evidence has been given tending to show that 
the prophylactic was polluted during the opening of the 
bottie. On what grounds, then, were the laboratory and 
its director indicted ? Even if the bacillus had entered 
during the complicated process of manufacture, the blame 
can hardly be attached io the director, who cannot him- 
self superintend the preparation of each bottle. As for 
the omission of the carbolic acid, the inventor of the 
prophylactic was himself surely the best judge of how it 
was to be made. 
The serious part of the affair seems to lie, not so much 
in the loss of life due to the accident itself, considerable 
as that was, but in the much greater loss which probably 
followed the suspicion thrown upon the prophylactic by 
the apparently erroneous judgment of the commission, and, 
more even than this, in a certain ingratitude shown in 
India to a man who is one of the very greatest benefactors 
it has ever had. Haffkine not only elaborated the method 
of immunisation by dead culture, but, where many a man 
of science would have contented himself with merely 
writing an article on the subject, he addressed himself, on 
the contrary, to the much more difficult practical verifi- 
cation. I well remember when he arrived in India with 
his anti-cholera vaccine and by his energy and perseverance 
gradually forced his ideas upon the people and the Govern- 
ment. When the frightful calamity of the plague over- 
took the country in 1896, largely, my opinion, owing 
to the inadequacy of the sanitary organisation and to 
want of firmness and resolution in the authorities, when 
measure after measure failed and the people were dying 
by hundreds of thousands, then Haffkine was the only 
one who made any successful stand at all against the 
storm. Quickly inventing his anti-plague prophylactic and 
forcing the authorities along with him, though he could 
not control the disaster, he at least checked it by saving 
thousands, if not hundreds of thousands, of human beings, 
who now owe their lives solely to him. The fact that 
more than six million doses of the prophylactic have been 
issued in India alone attests the success and magnitude of 
his work. Yet he has received for it less than nothing. 
For services which compared with his are really of a 
trifling nature, all kinds of officials receive in many cases 
pensions, promotion, and decorations. As for him, not 
only has he received no adequate recognition for his 
immense service, but he has been blamed for an accident 
good 
Great 
was 
labor- 
in 
which could not have been due to his fault, and it is 
doubtful whether he will ever return to a country which 
has treated him—I can only say—so ungratefully. Con- 
templating this history, we cannot help being filled with 
indignation at it. India seems to be becoming quite 
notorious for its treatment of scientific workers, susgest- 
ing ignorance both of science and of the importance of 
science. I remember the persecution suffered by Colonel 
King as the result of his work on vaccination, the com- 
plete want of gratitude shown to Mr. Hankin for his 
sreat work on the prevention of cholera, and_ several 
similar cases. While all kinds of people climb easily into 
the seats of honour, it seems that the men of real merit 
are fortunate if only they can escape without censure. 
I think T shall be excused for writing somewhat strongly 
en a subject on which T have long felt still more strongly, 
and on which I have reason to know many others feel as 
strongly as myself without being as free as T am to express 
NO. 1951, VOL. 75] 
their opinions. It appears to me a foolish thing for a 
nation to treat great men as we have sometimes treated 
ours, and the case of Mr. Haffkine—to whom, as he is 
a foreigner, we are doubly bound to show national grati- 
tude—seems to be a glaring example of such treatment. 
I hope that steps will be taken to press upon the India 
Office the need for a reconsideration of the affair; the 
reputation of the whole country is concerned in it. 
March 19. Roxatp Ross. 
Mean or Median. 
Tue article by Mr. Francis Galton in your issue of 
March 7, entitled *‘ Vox Populi,’’ is exceedingly interest- 
ing, and the variations in the estimates of individual com- 
petitors afford an admirable instance of the advantage to 
be derived from the use of the weighbridge at live-stock 
markets in preference to buyers and sellers relying on their 
own judgments; but the letter raises several interesting 
points as to the theoretical treatment of statistical data, 
to two of which | should like to allude. 
In the first place, as to bias. No doubt, in estimating 
carcase weights in such a competition as that referred to 
by Mr. Galton, each competitor judges as truly as he can. 
But has a butcher (buyer) had his judgment to any extent 
warped in the course of years through having constantly 
had to judge of the weight of a beast (when buying) so as 
to be on the safe side, and secure himself from loss in the 
event of its not cutting up so well as he anticipated? If 
so, it might be expected that buyers would have an in- 
slinctive tendency to under-estimate the weights of animals ; 
and similarly farmers (sellers) might be expected to over- 
estimate. This tendency, on either side, should, of course, 
not be large, as constant intercourse between ‘buyers and 
sellers has raised such transactions almost to the point of 
a fine art. I should therefore like to ask Mr. Galton 
whether he has any information showing the proportion 
of these 787 competitors who were farmers and butchers re- 
spectively. It is very interesting to observe, from the 
figures given, that the estimated weights at each decile are 
throughout the whole series invariably below the weights 
which might be anticipated from the normal Jaw of error. 
This rather looks as if buyers were in a majority in this 
competition: a not impossible suggestion, since, although 
farmers dcubtless attend such exhibitions in larger numbers 
than butchers, yet the latter would, in a weight-judging 
competition, probably be more numerous than the former, 
at least relatively, if not actually. 
The second, and more important, point to which I desire 
to direct attention is the use of the median in this con- 
nection, and I could wish that Mr. Galton had also calcu- 
lated the arithmetic mean of the 787 observations. I 
should, in fact, like to strike a note of hesitation in regard 
to the too general use of the median in preference to the 
mean. ‘The former has several advantages, one of which 
is that it is a form of ‘‘ average’? which can be very 
readily calculated. It is also very useful in cases such 
as those referred to in Mr. Galton’s letter in NATURE 
of the preceding week, where it is desirable to eliminate 
one or two ‘‘cranks’’ whose opinion might have undue 
weight among a relatively small number of other opinions— 
in cases, in fact, where the distribution of opinions is 
known to be very erratic. But is this the case here? I 
am not sure that Mr. Galton is quite right in regarding 
the present instance as a case of ‘* vox populi’’ at all. It 
is to be remembered that the great bullk of the trade in 
English cattle--and consequently the determination of the 
price of our native bee{—is the result of transactions such 
as the competition in question is intended to test. Cattle 
are practically sold by inspection, and the judgment of 
buyer and seller as to how much beef there is in a given 
ox is really much more a matter of skill than of popular 
judgnient; their livelihood depends upon the accuracy of 
such judgments. In such circumstances, is the median a 
nearer approximation to the truth than the mean? Here 
the question could be answered by calculating the arith- 
metic mean. I have not the actual figures, but judging 
from the data in Mr. Galton’s article, the mean would 
scem to be approximately 1196 lb., which is much closer 
to the ascertained weight (1198 Ib.) than the medica 
(H2070 1b.) 
