Land & Water 
August 2 2, .19 18 
the one side than on tlu- other, tlie effect is tliat what may hi- 
called the organism of the reserve is more weakened on the one 
side than on the other, and this result is not exactly dependent 
upon the rate of individual loss. 
EXHAUSTION OF RESERVES AND ITS EFFECTS 
Let mi give an example ; it is an over simple example, 
and one that could not happen in practice, but it illustrates 
the; point. Two commanders have each fifty divisions, of 
which they pit twenty each one against the other, each 
keeping thirty in reserve. Before any of these thirty is 
moved upon either side into the battle you have a situation 
in which each side has what may be called twenty wounded 
divisions and thirty unwounded or fresh ones. Let us 
suppose the individual losses in the wounded divisions on 
both sides to have been the same. The battle takes such a 
form that the first commander is compelled within the space 
of, say, four days to put in twenty of his thirty reserve divi- 
sions. Some of them do not suffer very heavily, but they 
have all been through the mill, and have lost, say, a quarter 
of their effectives. The first commander, however, has only 
had to put into the battle during the same space of time 
fifteen divisions. It is true that their aggregate loss is as 
great as his opponent's, because they have lost not a quarter, 
but a third of their effectives at the end of the period. Here 
vou have a situation in which at the particular stage we 
are examining, losses have been equal on both sides as counted 
in individuals. But losses as counted in damaged divisions 
have been much heavier on the one side than on the other. 
For at the stage which we are examining the one commander 
has fifteen completely fresh divisions, the other only ten. 
The same process continues. All the ten divisions remain- 
ing to the less fortunate commander are thrown into the 
battle in the course of, say, another three days, while his 
opponent is compelled to put in only eight. The rate of his 
opponent's losses are slightly higher, and therefore the total 
individual casualties on the two sides are equal. Here you 
have a phase of the battle in which the losses on the two 
sides are still equal, and yet the one party has seven complete 
divisions perfectly fresh which he can throw into the battle, 
and with which he can decide the issue, while the other has 
no fresh divisions left at all. 
That is what one means by saying that reserves may be 
exhausted in a proportion higher than the proportion of 
total losses. And that is why men note not only total appro.xi- 
mate casualties on both sides, but the rate at which divisions 
are used as well. 
A division having been put into an action is withdrawn 
when it has suffered a certain proportion of loss ; the pro- 
portion is determined by the judgment of the general com- 
manding the whole group upon the way in whicji he should 
use his power. It may be withdrawn after comparatively 
slight loss, or only after very heavy loss. At any rate, when 
it is withdrawn it has to be rested, recruited, reorganised 
lo some extent, and cannot appear in the field again until 
after an interval of time 'which varies ;- which has some- 
times under extreme pressure (as during the last great German 
effort in March and April) been only a week or two, but which 
on the average must be much longer and in which a delay 
with the minimum measured in weeks and the maximum 
measured in months, is the decisive feature in the whole 
business. It is a process of rotation working within necessary 
limits. Some months ago I compared it in these columns 
(borrowing the metaphor from another writer) to the use of 
a number of tractors upon a road. One tractor has to go 
back for repairs after losing a comparatively small amount 
of its parts, another goes back after losing more. But once 
you get your tractor back out of use into the repairing shops 
it cannot be returned for some little, time. 
COMPARISON OF RESULTS 
It is clear, then, that the exhaustion of enemy reserves, 
as measured in divisions and as compared with your own 
exhaustion on the same standard, is a capital element in 
the comparison of effective power, quite as important as 
the comparison of total numbers and their rate of loss. 
Now, if we examine the results of these two great battles, 
the Second Battle of the Tardenois or of the Marne, and the 
recent Battle of Amiens or Third Battle of the Somme, 
which is not vet wholly concluded, we shall find that, apart 
•from a recovery of the initiative there has been as the final 
•effect of these two enormous and rapid actions, a reversal 
in the superiority of effective power. 
When the enemy launched his great offensive, which was 
also to be his decisive offensive on Monda\', July 15th, he 
still had superiorit}' in numbers, and he probably had a 
sup-riority in the numbcl■^ of his reserve divisions. After 
allowing for those which were to be put immediately into 
action and for those necessary to hold other ])arts of the 
line, there would seem to ha\'e been about sixty divisions 
fresh and vet marked as reserve divisions to be tlirown in 
during the course of succeeding actions throughout the 
fighting season. Such a form was given to the battle between 
Soissons and Rheims after the Franco-American surprise on 
July i8th that the enemj''s reserve divisions began to be 
sucked in at a perilously rapid rate. It was necessary to 
throw a great number in rapidlj- to avert disaster, but the 
pace was increased by the hesitating policy of the enemy, 
who could not' make up his mind at first whether to retire 
at once or to hold to the furthest possible lines. By the end 
of that battle — that is, by the enemy retirement beyond the 
Vesle — he had certainly exhausted his original reserves more 
than had the Allies. Then came the second blow in front 
of Amiens on August 8th, and in a few days . the process 
was repeated. The insufficient remainder of his reserve was 
rapidly drawn upon at a much more rapid rate than the 
large reserve of the Allies, and the process begun three weeks 
before was heavily accentuated. It was the estimate of a 
high authority — the highest authority we have in the matter 
in this- country — that by Thursday, the 15th — at the end, 
that is, of the first week of the Third Battle of the Somme — 
all but three-quarters of the enemy's original reserve had 
been used. Of sixty divisions, sixteen at the most remained, 
and since then a certain number more must have been put 
in to replace and relieve divisions withdrawn. And the 
margin is getting extremely small. Nothing comparable to 
such an exhaustion has taken place upon the Allied side. 
Now, if we turn from this side of exhaustion — the ex- 
haustion of reserve divisions — and consider the other element, 
the actual losses and the actual recruitment, we come to a 
similar conclusion. 
When the enemy launched his great and (for him) disastrous 
offensive upon Monday,' July 15th, he still had a clear 
numerical preponderance. The Americans were coming in 
at a certain rate, but if he had succeeded and inflicted upon 
us — as his success would have done — far greater losses than 
we inflicted upon him, his numerical preponderance would 
have increased rapidly. As a fact, he failed, and with the 
recovery of the initiative by the Allies their first great captures 
of prisoners and material, coupled with the steady arri\-al 
of the American recruitment in the field, the tide began to 
turn. Whether numerical superiority was attained by the 
end of the Soissons -Rheims battle I do not know. It has 
certainly been attained to-day ; and American recruitment 
comes in steadily at a rate which more than makes up for 
the actual rate of loss — apart from the fact that the enemy's 
rate of loss, allowing for the n-ery light casualties at the 
beginning of the present action — must be far greater than 
ours. 
ESTIMATE OF LOSSES 
It is a rough but not an inaccurate estimate to put his 
total losses in the past month of defeat at more than 350,000 
and less than 400,000. Of these the- prisoners probably 
represent about one-fifth ; more than 70,000, but less than 
80,000 ; the guns he has lost, by the way, arc over 1,700. ■ 
Of the rema.inlng four-fifths, two-fifths at least are either 
killed or are definitive losses, because they will not return 
to active service. The remaining two-fifths will come back 
after an average of about four months. These and past 
hospital cases with class 19EO, which is about to appear in 
the field, and which numbers about 450,000 lads, are the 
whole of the recruitment he has in sight betw:een this and 
the latter end of next year's fighting season. Another way 
of putting it is to say that his immediate losses in this last 
month alone are nearly equal to his recruitment in new men 
for the next nine or twelve months, and that his positive 
losses — that is, those who will never return — were in that 
one month at least half such recruitment. 
Meanwhile, on the Allied side the whole of the French 
1920 class is kept back until next year, and the American 
recruitment pours in at a rate which puts men of far better 
physique and far better mihtary age into the field at a rate 
at least four times that of the enemy recruitment. 
■K » « , . 
That is how the strategic situation now stands, and that is 
what we mean when we say that the Allied marshals have 
recovered the initiative at the beginning of these operations, 
and have since also recovered superiority of effective power 
in both its branches, total numbers, and numbers of fresh 
units in reserve. It is difficult to see how these advantages, 
now definitely obtained, can be lost through any military 
cause. Thev would seem to' be final. 
