Aware that a prominent weak point in M. Estrade's design is that, no matter what size we make cylinders and wheels, we have ultimately to depend on the boiler for power, M. Nansouty argues that M. Estrade having provided more surface than is to be found in any other engine, must be successful. But the total heating surface in the engine, which we illustrate, is but 1,400 square feet, while that of the Great Western engines, on which he lays such stress, is 2,300 square feet, and the table which he gives of the heating surface of various English engines really means very little. It is quite true that there are no engines working in England with much over 1,500 square feet of surface, except those on the broad gauge, but it does not follow that because they manage to make an average of 53 miles an hour that an addition of 500 square feet would enable them to run at a speed higher by 20 miles an hour. There are engines in France, however, which have as much as 1,600 square feet, as, for example, on the Paris-Orleans line, but we have never heard that these engines attain a speed of 80 miles an hour. Leaving the question of boiler power, M. Nansouty goes on to consider the question of adhesion. About this he says: Is the locomotive proposed by M. Estrade under abnormal conditions as to weight and adhesion? This appears to have been doubted, especially taking into consideration its height and elegant appearance. We shall again reply here by figures, while remarking that the adhesion of locomotives increases with the speed, according to laws still unknown or imperfectly understood, and that consequently for extreme speeds, ignorance of the value of the coefficiency of adhesion f in the formula d 2 I fP = 0.65 p ------- - R D renders it impossible to pronounce upon it before the trials earnestly and justly demanded by the author of this new system. In present practice f = 1/7 is admitted. M. Nansouty gives in a table a _resume_ of the experience on this subject, and goes on: