s 
LAND & WATER 
August" 9, 1917 
accession of'strength from this quarter is closed. 
So much for the statement of enemy strength. We always 
specialise upon that of the German Kmi^ire. lirst because 
its proportion to the total belligerent force against us is known 
(it is one-half) and therefore it is an exact ciiterion ; secondly 
because our information upon it is so much more extensive 
and accurate than our information upon any other section of 
the forces opposed to us. ; 
The other limb is the rate of loss, and the bridge uniting 
the two is the estimate of men in depot at any moment, and 
coming in as hospital returns for recruitment month by month 
as the war proceeds. I- 
The estimate of enemy losses in any one action or over 
any recent and short period of time, is largely a matter of 
guess-work, and the tendency is nearly always to exaggerate. 
The only positive evidence available at first is that afforded 
by prisoners and perhaps a few documents —captured on the 
field. Then,.; shortly afterwards, comes the tabulation of 
divisions withdrawn, coupled witli our knowledge of the average 
loss which a division is expected to stand before it is with- 
drawn. Later comes our knowledge through other forms 
of intelligence ; lists printed within the enemy countries, 
newspaper reports, corroborative evidehce of other kinds 
(much of which is necessarily secret) and at last, after a con- 
siderable period of time, the process of calculation gets nar- 
rowed to within comparatively small margins of error. 
For instance, we know now to within a very small margin 
of error what the enemy lost in his five months of effort against 
the Verdun sector. It was, with less than 10 per cent, of 
error either way (and probably within 5 per cent) a loss of 
Somewhat over 100,000 a month, the total losses being in the 
aeighbourhood of 520,000 or rather more.' To put it another 
way : it would surprise all those very many men whose busi- 
ness it has been to collect and co-ordinate evidence upon the 
matter, if some years after the war (supposing full information 
to be then available) it was found that the total losses were as 
much as 550,000 (excluding, of course; sickness and accident), 
and if the numbers were given as low as 506,000 we could i)rove 
positively that there was concealment: In the, same way, 
after a sufficient time for the gathering of -much evidence, 
it is estimated that his losses during the five months of the 
Somme were in tlie neighbourhood of; yoo.ooo, though this 
figure, being more recent, is perhaps subject to a slightly larger 
margin of error. . ■ ' 
When we come to general losses over a very great period 
of time we can get still more accurate results. We are more 
certain, for instance, of such an estimate ;as' appeared in these 
columns in the early part of iqi'J, giving the German losses 
up to the end of December before, thariwe are of the losses 
in any particular action or set of actions.": ; . :,. 
This category, then, of general lossesi. is. a dependable 
one within a great degree of accuracy,, though that degree 
lowers in proportion as the events witliwhidr we are dealing 
are recent. " •'V'^ ^ 
The second element in this limb is the^ fate of hospital 
returns. Of general losses, a very great number return to 
sotm kind of service, and after the experience of so many months 
of war this proportion is known. It is a little over 60 per cent. 
Of 100 men who appear in a total list of, casualties and sick, 
some 40 or perhaps a little less, but not much less, are out of 
the war for good. 
That phrase " any capacity " is, of course, liable to mis- 
construction. There is nothing to prevent your keeping a 
hopelessly mutilated man, or a man. whose health is perma- 
nently injured, in uniform and upon the ration strength, and 
you can, if you choose, give him some sort of work to do 
which is of little service to the armv, but which gives you some 
excuse for not destroying him. Political considerations here 
come in and vary the proportion so retained. 
But though this introduces an clement of error in the paper 
strength of your opponent, the error is of no practical con- 
sequence. WTiether, for instance, a man unable to take any 
active exercise, but helping to keep accounts in a hospital, is 
on the ration strength or is employed as a civilian, is, for the 
purposes of inihtary calculation, quite imimportant. The real 
interest lies in an estimate of the proportion that can he used 
again in active service and here the calculation is subject 
to considerable difficulties. But we have certain rules. 
In the first place, it is clear that the real interest again 
lies in the numbers who can return to the same active service 
as that which thcv were discharging before, not to auxiliary 
ser\'ice of any kind, nor even to services less exacting than those 
they first performed, but to the very same service. P'or in- 
stance, a man may have been once in the units which in rota- 
tion hold the front trenches. He suffered severe illness or 
mutilation which forbid him to be so used again, but he may 
none the less be used for services which are of an active military 
sort. He may be attached to transport or even to certain 
services in coanection with the heavy artillery behind the 
lines, such makeshifts do not fill the gaps. The fighting line 
depends on the recruitme;it of its infantry actively engaged. 
Another rule %chich we must bear in mind in connection 
with this difticulty is that the proportion of men returnable 
has, paradoxically enough, increased with the duration of the 
war. When the strain was less severe a man suffering from 
sucii and such a disability would be discharged or put on to easy 
auxiliary work ; to-day he will be forced into harder work 
and even perhaps into the original active service from which 
he was evacuated as sick or wounded. 
Roughly speaking, one may say that, of a total casualty 
list, while some 60 odd per cent, return to some sort of service, 
little more than 40 per cent, returned in the earlier part of 
the war to exactly the same service as they fulfilled before, 
and even to-day not more than 50 per cent. 
For instance, when we said in these columns last January 
that the Germans " had in sight " for the fighting season of 
1917, up to some date in August, about a million, of whom 
half a million were already in the depots, it was allowing for 
hospital returns which may have been wrong by say, 7 per 
cent, or S per cent. But as these hospital returns will form 
less than half the whole, the total margin of error was only one 
of 3. per cent., or 4 per cent., the rest being the last of the 
combings out who had been put into depots and the class 1918, 
the numbers of which were accurately known. 
It will be seen from the above that this kind of calculation, 
even where it is applied to los,ses which are less certain than 
establishment and total man-power, is sufficiently accurate 
for all practical purposes of judgment. It is not as absolute 
as an honest balance-sheet in business or a census return, 
wherein there is virtually no margin of error at all, but it is 
far more accurate than, let us say, the estimate of the rate at 
which a piece of railway transport can be conducted, or the 
probable mortality rate of a great city in some forthcoming 
period. To give a parallel instance which will be familiar 
to everyone ; It is not as accurate as the Government forecast 
of a harvest made a week or two before the grain is gathered, 
but it is much more accurate than such a forecast made at the 
end of June. 
Proofs of .\ccuracy. 
Finally, we must, note that the accuracy of such estimates 
is constantly checked and corroborated by the event. 
For instance, evidence of a hundred different kinds co- 
ordinated results in this estimate of German losses alone to 
the end of 1915, after the first 17 months of war. Losses in 
death, counting all those who have died that were ever on 
the ration strength, about one million ; definitive losses, 
that is, losses in dead and prisoners and hopelessly sick and 
mutilated, just over two millions ; losses including those who 
cannot go. back to the same active service as before, a good 
deal--say, a quarter — over three million. It will be remem- 
bered that this was very fully gone into in the articles appear- 
ing in Land & W.\ter in March 1916. 
Turn to, the estimates of the second 17 months, that is, up 
to the beginning of June of this year, and you find all these 
.figures increased, by somewhat over 75 per cent . , but not 80 
per cent. You are less certain of the last 17 months than of 
the first, because the time is more recent, but within that 
margin of error you are corroborated by the nature of the 
fighting. German wastage has not gone on at the same rate 
because of the long months during which no pressure was 
exercised upon the enemy between the Baltic and the Black 
Sea, where the Germans had one-third of their army.' 
The estimates are again checked and corroborated by the 
rate at which the new classes are called up. Each such 
summons has come accurately within two or at most three 
weeks to the time-table suggested by the estimates. 
Thus, if the losses had been what they were presumed to be, 
class 1918 in Germany ought to have been called up at the 
end of last j-ear, and class 1919 not earlier than the end of 
April and not later than the beginning of June of this year. 
The known events in both cases corroborated the estimates 
previously made. 
We may take it, then, that these general estimates appear- 
ing from time to time of enemy (and particularly of German), 
strength- -which strength is a function of (a) total man- 
power, (b) losses, and (c) recruitment — are accurate for all 
practical purposes. The margin of error is not what you 
would get in most business estimates, and the totals are work- 
ing hypotheses so near to truth that they correspond to the 
results that should flow from them. 
But even if they have this accuracy, are they worth mak- 
mg } Are they of value to our judgment of the war and 
therefore to the maintenance of civilian tenacity and sound 
sense during the strain which the nation is undergoing } That 
is the second part of the proposition with which I shall deal 
next week. h. Belloc 
