grounds. It is certain that General Thomas has not consciously laid claim to an idea which did not possess him; but I apprehend that developed facts have been so mingled in his mind with impression coincident in some particulars,though essentially different in the material points, that his memory has failed to draw the exact distinction between them. It is, however, due to him to say that the idea may have been in his mind that Bragg might cross the mountain to Sparta, and that he did not distinctly express it me to, imagining that I also entertained it myself.

But I do not propose to draw any advantage from the question whether or not a proposition was made to me to concentrate at Sparta. If it had been made, I should have judged it according to its merits with rejected it on grounds which I will state.

Besides the road which crossed from Jasper to Decherd and the one which ascends the valley and thence goes to Crossville there are no less than three roads by which the enemy could ascend the mountain to debouch from the Sequatchie Valley: First, the Therman road, which passers through or near Altamont, and then branches into at least four roads that descend the mountain into the plains of Middle Tennessee between Decherd and McMinnville, a distance of about 40 miles; second, a road which ascends the mountain at Dunlap and passes to McMinnville; third a road which ascends the mountain a short distance below Pikeville and branches on the mountain, the left-had branch going to McMinnville and the right hand forking again some 20 miles from McMinnville, one fork going to the latter place and the other to Sparta. There is also a road on the top of the mountain connecting all these roads. These geographical features would enable the enemy to arrive within 20 miles of McMinnville by not less than roads before determining whether he would move on that point or Sparta, and by covering his movements with his superior cavalry force he could easily arrive within 6 or 8 miles of either of those points before his destination could be know at all, and it is 22 miles at least from McMinnville to Sparta. If I had been at Sparta he could have been at McMinnvile and in possession of my line of supplies before I could have known it. If I awaited at McMinnville the development of his plan he could have gone to Sparta and pursued his course as he did. If I had divided my force between McMinnville and Sparta, to anticipate him at both points, he could have advanced with reasonable probability of success against either of them; and if the fractions should have been so strongly fortified as not to warrant an attack, he could have avoided them, thrown himself between the two, and thus have forced to retreat separately, or attempt the offensive against a concentrated force. General Thomas' own experience at McMinnville in obtaining information on which success would have depended confirms my answer to one phase of this proposition and is applicable to all of them.

On the 31st he reports:

The general impressions is that the enemy is advancing, but I have yet to see the person who has seen any of the Chattanooga forces proper.

And on the 2nd of September, in reply to the discretional instructions heretofore alluded to, he says:

I will start to-morrow. I have heard again that the enemy intends advancing on this place by the Therman, Dunlap, and sparta roads. By concentrating at Murfreesborough we shall be within striking distance of this place. By convenient roads our main force can be thrown upon the enemy between this and Decherd or Hillsborough, overcome him, and drive him toward Sparta, his longest line of retreat. A