398 
PROFESSOR 0. REYNOLDS AND MR. W. H. MOORBY 
These two large masses of iron had hitherto been left on the brake, but it seemed 
probable that they w^ould very much affect the flow of heat away from it between 
any pair of consecutive trials (Part I., par. 33), for they continued to rise in 
temperature during the whole of any day on wdiich experiments were made, and 
evidently they would absorb heat more rapidl}^ when cold in the early part of the 
day than when hot later. It was therefore decided to remove them. Their combined 
moment about the engine shaft was — 44‘i2 ft.-lbs. 
No allowance was made for this alteration in the loading of the brake, and, 
consequently, the moment in these trials was 1244T2 ft.-lbs., this figure having 
been used in the calculations given. 
In order to bring the trials under some general denomination, this series has not 
been further reduced, nor combined with a corresponding set of light trials. 
With the intention of stopping the leakage at the bottom cock, I had had some 
more packing placed in the gland surrounding the cock spindle. This did, to some 
extent, reduce the leakage, but it also had another effect which wdll be referred to 
under Series V, 
Series V, 
40. For the purpose of keeping the loads on the brake at the values carried by 
trials preceding the removal of the rider and balance weights, one of the 25 lb. 
hanger weights was removed, and for it were substituted some lead sheets w^eighing 
13-97 lbs. 
Ibis lead weight then corresponded with the initial want of balance to a moment 
of 100 foot lbs., made up as follows :— 
Want of balance.44-12 foot lbs. 
Moment of lead weight . . 55-88 ,, 
After these trials had been made, I determined, with Professor IIeynolds, by 
means of a spring balance, the force necessary to move the bottom cock. This was 
found to amount to a moment of 30 ft.-lbs. on the brake, and on this account this 
series of trials, though appearing in the final tables, have not been allowed any 
weight in the calculation of the final mean value of K. The preliminaiy reduction 
of Table A gave what were apparently very good values of K, but this only shows 
the small effect on the mean moment produced by variations in the resistance 
offered to the brake’s motion, and this although its period of oscillation was very 
long. 
Series VI. 
41. These trials differ from those of Series V. only in the fact that the extra 
