280 Col. Bemifoy on the Resistance of Water, [April, 



foot in diameter ; at first sight it miglit appear that the largex 

 the surface, tlie more accurate the experiment, which, though 

 true in theory, is false in practice, lor large sohds both by bulk 

 and weight became unwieldy, and consequently difficult to 

 manage ; and when experiments were to be made beneath the 

 surface of the water, considerable trouble occurred in accurately 

 placing, and firmly securing, the immersed body to the conduc- 

 tor. When practicable, only one bar should be used in attach- 

 ing the upper and lower bodies to each other ; the shape of the 

 bar should be elliptical with the transverse or longer diameter 

 parallel to the centre hne of the experimented body. This form 

 of bar is advantageous on many accounts ; it meets with less 

 resistance, is not so liable to bend from the impulse of the water, 

 and answers the purpose of a rudder by a small alteration in its 

 parallelism, which, without affecting the accuracy of the experi- 

 ment, will prevent the conductor from deviating with its attached 

 solid from the intended course through the water. It would be 

 a further improvement if the tremulous motion of the iron bar 

 caused by its elasticity was prevented, which might be done by 

 placing the bar in a metal case, and filling up the vacuity with 

 melted lead. The part of the bar which passes through the con- 

 ductor should be circular, that the figure attached may be 

 placed accurately by turning the bar ; and a mark should be 

 made on the bar, and another on the upper part of the conduc- 

 tor, which ought to correspond when both are truly adjusted. 

 At the bottom of the bar, a concave screw is cut, which fits into 

 one of a contrary description that projects from the solid whose 

 resistance is to be found. To prevent the bar from sinking, and 

 being lost when necessary to detach it, the upper part was 

 formed into a hook, and to this was fastened a line. 



One of the xprincipal reasons for recommending one bar in 

 preference to two, originated from the loss of time, trouble, and 

 vexation, in the year 1796, proceeding from the use of two. 

 Many of the experiments made at that time were so extremely 

 discordant as to induce a belief that the particles of water when 

 once displaced did not arrange themselves in the same manner ; 

 but on reflection it was thought that the bars which were cylin- 

 drical, one inch and a half in diameter, and six feet asunder, were 

 sufficiently close for the eddy water of the foremost to affect the 

 resistance of the hindermost, which, on a further separation of 

 the bars to nine feet, was found to be the case ; at the same time 

 it was deemed adviseable to alter the shape from a circular to an 

 angular figure. In making experiments of this kind, it is neces- 

 sary to have a considerable length of line to draw the figures by. 

 To avoid the inconvenience of a high mast which, if a singlo 

 pulley was used, would be unavoidable. A system of pulleys 

 was adopted, and it was fcmnd from experience that two double 

 blocks answered the purpose exceedingly well. The sheaves or 

 whueb of the block were not plucodin one sl^ell, or side l»y 



