,August It), 1917 
LAND & WATER 
wooded Carpathian range from the DornaW'atra corner to the 
neighbourhood of Focsani. 
So long as Czernowitz stood, numerous roads and a whole 
system of short local railways sers'ed thafcentre and turned it 
iiito a place of supply for everything to the south. With 
Czernowitz gone the main Sereth road and railway become 
almost useless. They ser\-e only to contribute supplies that 
may come through Jassy, and jassy only communicates with 
the Russian supply centres based on Kiev in the most 
devious manner through Odessa. 
At the same time the Russian al)andonment of the Bukovina 
(of which Cjternowitz is the capital) turns the whole Allied line 
in Moldavia into a salient, the obvious policy for reducing 
which (and driving the Russo-Roumanians out of what is left 
of Roumania) is to attack the two wings and at the same time 
tu bring pressure on the critical pillar point of the Dorna Watra 
comer. It is upon this accoimt that Mackensen is attacking 
as hard as he can northward from l"ocsani. Every mile of 
advance he makes, and every mile of retirement suffered by the 
Russians in the Bukovina to the north, accentuates the salient 
and makes it more and more difhcult to hold Mofdavia. The 
Roumanian Court has already left Jassy, and the pillar point 
of the Dorna Watra comer is yielding. At the moment of 
writing (Tuesday morning) the last news is that Mackensen 
has reached Pincin and immediately threatens the junction 
between the two parallel railway lioes'that serve the Molda- 
vian plain. 
It need hardly be added that the whole of this unfortunate 
situation in Northern Roumania is ultimately.due to the break- 
down of the nth Russian army in front of Tarnopol (now 
nearly a month old). 
So true is it that the strength of an army depends on its 
cohesion, that is, on the validity of every part, and that this in 
its turn depends upon universal discipline. \\ hether Moldavia 
will be saved or no only the event can show ; but so much is 
certain : That. only an unexpected Russian resistance in the 
so M^lIOO 
north can save it. Strategically, if the retirement through 
the Bukovina continues beyond the frontier, Moldavia has 
gone. 
The Importance of Calculation— II 
The Advantage 
I think I showed clearly enough last week within how close a 
dffgrcc of accuracy we could come in these figures of belligerent 
• numbers, and how legitimate was- the use of such estimates 
merely in the point of exactitude. But doubts upon the mere 
reliability of such statements (which can onlv arise from 
Jiastcjn reading or from imjierfect appreciation of the sources) 
is only ono" half of the trouble. The other, and perhaps the 
larger half, is the feeling that, accuratej&s the calculations may 
bt under given conditions, those conditions change so 
abruptly and their change is itself so incalculable that time 
. and trouble are wasted in attending to any calculation at all. 
Here is an example of what I iticiof : 
In the autumn of 1915 there was published in these columns 
a verycareful estimate of the enemy's numerical position. 
The size of the .'Kustro-Hungariiii field army wks slightly over- 
estimated from a lack, at that time,' of sufficient evidence ; 
but that of the (ierman field arnrv and its losses were given 
• accurately enough because a suflficient time had elapsed to 
correct the errors made in earher Calculations at the beginning 
: of the war. 
In estimating the resen-es "available for filling gaps the 
estimates made at that moment allowed for a certain large 
nuoibcr of civilians of military age as being necessary to the 
"'mnningof-thecountry." Up to that moment this large num- 
Ix-r had been kept back in civilian employment. It was a 
matter of necessity that they shou}d be so kept back ; and the 
amount of available reser\e man*powerwas by that extent 
reduced. But shortly after this moment the enemv originated a 
totally novel practice, hitherto unknown in civilised war, and 
therefore not allowed for in any such estimates. The (lennans, 
began, at first timidly, afterwards more boldly, to enslave the 
populations of occupied countries. -^ It was clear that as this 
no\el iw)licy proceeded and developed, very large number?! of 
those hitherto eliminated frornthe estimated reserve of man- 
power as necessary- for civilian employments, would be 
released for the army. As artiatter of fact the power of 
" combing out " such men continued during the whole of 
I()i0 and only came to its final limit at the beginning of the 
present j'ear. Therefore, the estimates made in the autumn 
of 1915, though perfectly accurate for their time and 
circumstance, needed sharp revision in favour of the enemy 
within six months after the time they were made. 
Here is another example. A full statement on German 
numbers was made in L.\Nn & W.\ti;r early in ioi(>. referring 
to the various categories of German losses (numiiers of dead, 
number of definiti\e losses, number of losses from active 
service, etc.) up to the jist December, ifjiO, and it was there 
shown that the total number of dead off the gross ration 
strengti) was then approximately one million. The total 
definitive losses, therefore, a little over two million, and the 
losses to full active service some three million and a quarter. 
Now anyone reading that estimate (which events have proved 
to be perfectly accurate) might well have extended it when 
judging it in his own mind somewhat as follows : 
" These are the losses for the first 17 months of the 
war. We may take it that the same rate will continue, and 
if the war lasts as much as another 17 months we shall have 
double- that number of German dead ; double that number of 
German definitive losses, and .so on. We know the rate of 
German recruitment, and we see that under such circumstances 
the German lines as they now exist cannot be maintaiiied 
throughout the fighting season of 1917." 
No such conclusion was drawn, of course, in these coluitins. 
That would have been prophecy, and prophecy is a tljing 
which this paper, T am glacl to say, has never indulged' in. 
Indeed, a good deal of our space has been spe^tiih' wawting 
certain contemporaries ami their readers against pr6j)hec\ . 
whether optimistic or the reverse. The spirit that tallcec' 
of " the Russian Steam Roller " and " The Russians in Berlii 
by the autumn of 1914," has no more place here than the spiri' 
that talked of " a (ierman march on India " and of ending tlu 
war " still looking at Vimy Ridge." 
Nevertheless, it is not only conceivable but probable that 
many people would use exact and known statistics of the past 
first seventeen months of war, for the formation of a guess with 
regard to the future ; and when that future became the present 
their guess proved wrong. After another second seventeen 
months of war the German losses had not doubled. Thev 
had increased by little more than another three-quarters, so 
that there existed a sufficient though barelv sufficient margin 
of reserve, though not for the fighting season of 191 7. ' ,: 
Why was this ? What had happened in the inten-al }_ '•■ 
What had happened in the interval was the Russian break- 
down, with the consequent relief of the enemy from pretty 
well any sort of pressure over the whole of his Eastern front, 
and the corresponding rapid decline in his total casualty 
rate. 
Now it is easily comprehendible that with such experiences 
in mind (and I have only quoted two examples out of verv 
many that might have been chosen) a man might say : " These 
statements of numbers, however accurate for the time in which 
they are made, are a waste of energy and mislead the reader 
because other incalculable factors are perpetually coming in, 
and they make all the difference." 
I hope to persuade my readers that this is not the case. 
If you so misread the war as to imagine the numerical factor to 
