But Muhamed gives unmistakable signs of impatience to show that he has had enough of spelling. Thereupon, as a diversion and a reward, his kind master suggests the extraction of a few square and cubic roots. Muhamed appears delighted: these are his favourite problems: for he takes less interest than formerly in the most difficult multiplications and divisions. He doubtless thinks them beneath him. Krall therefore writes on the blackboard various numbers of which I did not take note. Moreover, as nobody now contests the fact that the horse works them with ease, it would hardly be interesting to reproduce here several rather grim problems of which numerous variants will be found in the accounts and reports of experiments signed by Drs. Mackenzie and Hartkopff, by Overbeck, Clarapede and many others. What strikes one particularly is the facility, the quickness, I was almost saying the joyous carelessness with which the strange mathematician gives the answers. The last figure is hardly chalked upon the board before the right hoof is striking off the units, followed immediately by the left hoof marking the tens. There is not a sign of attention or reflection; one is not even aware of the exact moment at which the horse looks at the problem: and the answer seems to spring automatically from an invisible intelligence. Mistakes are rare or frequent according as it happens to be a good or bad day with the horse; but, when he is told of them, he nearly always corrects them. Not unseldom, the number is reversed: 47, for instance, becomes 74; but he puts it right without demur when asked. I am manifestly dumbfounded; but perhaps these problems are prepared beforehand? If they were, it would be very extraordinary, but yet less surprising than their actual solution. Krall does not read this suspicion in my eyes, because they do not show it; nevertheless, to remove the least shade of it, he asks me to write a number of my own on the black-board for the horse to find the root.