Prof. J. E. Denton.-If I understand this paper, it seems to me inaccurate to claim that it records any measurement of a coefficient of friction which confirms the experiments on the friction of engines discussed at previous meetings. Mr. Webber has a table of coefficients in his paper which is the ratio of the power lost in friction to the power transmitted. But he includes the friction of the entire apparatus, embracing four bevel gears, two driving-pulleys, two spur-gears, and all the miscellaneous details. It is impossible to say that this ratio is the coefficient of friction on any one bearing, and it therefore does not confirm Mr. Woodbury's experiments in my opinion at all. I want very carefully to assert to-day that I do not believe there is a particle of proof in any investigation of friction ever made, that Morin's laws do not hold for ordinary practical oil-cups or restricted rates of feed. I have reviewed this point in the last volume of the Transactions in connection with Prof. Thurston's paper.

Prof. Wood.-I want to ask Prof. Denton one question. You speak in your remarks about the friction. Did you mean the friction or the coefficient?

Prof. Denton.-I mean that the coefficient of friction is practically constant.

Prof. Wood.-When you spoke of the shaft and the bearings, you meant the coefficient?

Prof. Denton.-Yes, sir.

Mr. C. J. H. Woodbury.-I find some reference has been made in the paper to former experiments of mine. I need only call attention to the fact that I read several years ago two papers before the Society, in which I expressly stated that the variation of the coefficient at different pressures was limited as a practical matter only to the smaller pressures which exist, especially in spinning machinery, where the pressure is so light and the film of oil so thick that the viscosity of the oil is an important part of the frictional resistance of the material.

Prof. Denton.-I would like to add a word in view of Mr. Woodbury's remarks, that in my discussion in bringing up this point I agreed with Mr. Woodbury entirely, that his experiments, just as he says, did apply to these particular conditions, and all that he claimed was thoroughly true; but that it does not fit Morin's conclusions. Morin worked at practical shafting fed with oil in the ordinary way.

Mr. Samuel Webber.*—I am not aware that the actual and ascertained facts of the paper, carefully noted during many days. with the assistance of a skillful practical mechanic, need any argument on my part to sustain them.

'Called out by Prof. Thurston's papers on the internal resistance of engines, they corroborate them fully in showing that a large increase of power was transmitted through the dynamometer without any perceptible increase in its internal friction, and that they agree thus with Mr. Woodbury in showing a great decrease in what I have called the coefficient of friction due to the load.

In a paper which I have in my possession, given me by Mr. Woodbury in 1883, he makes the coefficient of friction under a pressure of 1 pound per square inch .1700, under a pressure of 10 pounds per square inch .0348, under a pressure of 20 pounds per square inch .0257, and under a pressure of 40 pounds per square inch .0201, and it is these results which I have said my notes confirm.

I have not attacked Morin in any way, but may say that hav

* Author's closure under the Rules.

ing met him personally in early life, and knowing something of the conditions under which his experiments were made, I consider myself fully competent to judge as to the value of those experiments, and whether they limit the extent of human knowledge as to the laws of friction. I do not propose to say that the term "coefficient of friction," as I use it in the paper, is exact as to the friction of any one bearing or gear, but it was the only available term I could use to represent the proportion of the total of transmitted power absorbed in the transmitting instrument.

Whatever previous theories these facts may conflict with, they are facts, and as such are presented for record.

CCCXXXV.

CORNISH OR DOUBLE-BEAT PUMP VALVES.

BY A. F. NAGLE, CHICAGO, ILL.

(Member of the Society.)

In the earlier studies of the writer there were few problems which perplexed him more than to find a satisfactory theory for the construction of a double-beat pump valve. The text-books were silent on the subject, and engineering journals contained only illustrations without sufficient data for any analytical investigation. Hence he had to gather what facts could be found, and attempt to formulate a theory.

Valves were found with seats nearly an inch in width, sometimes flat and sometimes beveled. The unbalanced area was rarely determinable, and the weight was never given. The possible lift was frequently noted, but whether the valve ever reached this limit was not recorded. Sometimes valves were known to rise with such force as to break the stop provided, and then again they would seat with such violence as to endanger the safety of the pump. There seemed to be no intelligent practice on the subject, and the problem was evidently one which theory could not solve, and the only way was to make something, and then experiment with weights, springs, and air snifted or pumped in, until something passably good was arrived at.

The principal features into which the subject appeared to divide itself were:

1. The width of seats.

2. The unbalanced area.

3. Its weight.

4. Its lift.

5. Its form of body.

1. The width of seat. What should be the width of a valve seat? Theoretically, a knife edge, so that the same area should be presented to the water pressure before as after it is lifted. Practically, only sufficiently wide to sustain the pressure brought to