The plan under which nearly all routine work in connection with the General Staff, work that the C.I.G.S. would otherwise be obliged to concern himself with personally to a large extent, was delegated to a Deputy who was a Member of the Army Council was an admirable arrangement. It worked almost to perfection as far as I could see. It allowed Sir W. Robertson, in consultation with his Directors of Military Operations and of Intelligence, Generals Maurice and Macdonogh, to devote his attention to major questions embracing the conduct of the war on land as a whole. The Deputy in the meantime wrestled with the details, with the correspondence about points of secondary importance, in fact with the red tape if you like to call it that, while keeping in close and constant touch with the administrative departments and branches. Everybody advocates de-centralization in theory; Sir William actually carried it out in practice, reminding me of that Prince of military administrators, the late Sir H. Brackenbury. The Deputy's room opened off that of the C.I.G.S.; but on many days I never even saw him except when he looked in for a minute to ask if I had anything for him, or when I happened to walk home some part of the way to York House with him after the trouble was over for the day. It was intensely interesting to have the daily reports of casualties at the Western Front passing through one's hands, and to note the extent to which these mounted up on what might be called non-fighting days as compared to days of attack. As this was during the opening stages of the Flanders offensive subsequent to General Plumer's victory at Messines, these statistics were extremely instructive. I do not know whether the details have ever been worked out for the years 1915-17, but it looked to me at that time as if the losses in three weeks of ordinary trench-warfare came on the average to about the same total as did the losses in a regular formal assault of some section of the enemy's lines. Or, putting the thing in another form and supposing the above calculation to be correct, you would in three weeks of continuous attack in a given zone only lose the same number of men as you would lose in that same zone in a year of stagnant, unprofitable trench-warfare. Some of our offensives on the Western Front have been condemned on the grounds of their costliness in human life; but it has not been sufficiently realized in the country how heavy the losses were during periods of quiescence.