Page:The rise and fall of the Emperor Maximilian.djvu/190

 be the cares which are imposed upon me by my duty as commander-in-chief of the French army.

My letter of July 11 has set before your majesty the duties which fall upon me in connection with the approaching withdrawal of a considerable portion of the army under my command.

As a natural consequence of the apprehensions which I am justified in conceiving as to the part which the Mexican element plays in this country, I have the honour of acquainting your majesty that it will be impossible for me to leave any of my troops at Guaymas and at Mazatlan.

For a long time the Mexican government has been able, and ought, to have taken in hand the maintenance of the imperial power in these two places. I find myself compelled to leave La Sonora and Sinaloa to those resources alone which the government of your majesty has at its disposal, and I shall not delay in recalling the French troops who are occupying these far-off districts.

With regard to the officials who have given their support to your majesty's government, I take them to be too clever either to have compromised themselves fruitlessly, or to expose themselves to danger from events which they must have foreseen.

They have all of them understood hitherto, and I think will still understand, how to keep themselves clear from all danger.

Upon the whole, sire, I do not think that the evacuation of Monterey and Saltillo can produce the serious results for your country which you appear to dread.

In war, it is necessary to know how to make allowances for emergencies, and to be able to sacrifice a portion of territory in order to preserve the main part; and then, when the enemy is weakened by exhaustion or desertion, the offensive must again be resumed and victory established.

In order to arrive at this end, your majesty has and will continue to have at your disposal means (the Foreign Legion and the Austrian brigade) which will not leave you in embarrassment.

With the deepest respect, sire, &c.,

This letter—which, by the way, clearly shows the