Rule to see that water, by its own cause, cannot raise itself on high.

2 - f e - 2, 20, 200, 1000 - d 10 1, 10 1, 2 10 1 2/5 g - c b a - 1000 - h - 10 braccia rise. Here the water ...
The wheel a and the wheel b and c are of equal size, and wheel a has two parts of lever against 10, and wheel b has one part of counter-lever against 10 of lever, and likewise wheel c; besides this, wheel e has 20 buckets, each bucket filling with the 2 pounds of water that is between d and e. Therefore, when wheel c has made one turn, it will have consumed 40 pounds of water from the cistern e f; and when wheel a has made one turn, wheel c will have made 100 and will have consumed 4000 pounds of water from the cistern. Now the waters that are between g and h weigh 1000 pounds, so that, wheel a making a whole turn, it will pour into the cistern 500 pounds, because a whole turn of wheel a pours a quarter of all the buckets that are between g and h. Therefore the cistern will pour out 4000 pounds of water and will get back 500; and he who spends more than he gains, fails. The cistern thus pouring out 4000 pounds and in that same time getting back 500, you may make yourself certain that, as much as 500 is less than 4000, so much thinner will be the continuous vein of the water that rises than that which descends from the cistern; whence you will say that 500 goes into 4000 eight times. Therefore the water that rises will be eight times thinner than that which descends by this method.
And if you wish to make it rise higher, I assure you that, as much as it rises higher than the said height, so much thinner will it be, and it will behave like iron when it is drawn out, because between g and h there will always be 1000 pounds of water against 2 in the first motion.
Conclusion. I conclude that, as much as the water rises higher than the percussion of the water that lifts it, so much the more slender a conduit it will produce in continuous pouring. And if you did not want the water that moves the wheel to fill little boxes, like the French mills, it would be so much slower that, as to the resulting conduit, it would come to that same thing above.

Now one must consider the power of the blow, which is joined to the water that moves and not to the continuous conduit that results from it. Therefore the conduit of the water that results will be so much thinner than that which moves, as it is deprived of the blow. And even if you arrange that the water which moves remains attached to the descending part of the wheel, I doubt whether there is any advantage, because the water that falls upon the water standing still in the boxes does not make that impact which it would make in falling upon a hard thing; and because the water gives way to the blow, the blow is vain; and if you gain weight, the blow is lost; therefore the weight does as much as the blow.

This motion cannot continue, because the water of the cistern that gives the first motion must always weigh more than that drawn water which is poured into it. That principle of the lever which makes more way against its counter-lever will be the stronger.

1 - 3 - 4 - 100 - 3 - 5 - 200 - 19 - 190 - l