ON GRAPHIC METHODS IN MECHANICAL SCIENCE. 373 



Second Meport on the Development of Graphic Methods in Mecha- 

 nical iScience. By Professor H. S. Hele Shaw, M.Inst.G.E. 



[Ordered by the General Committee to be printed in extenso.'] 



Three years ago a preliminary report was presented on the above subject 

 by the Committee, consisting of Mr. W. H. Preece (Chairman), Mr. (now 

 Sir) Benjamin Baker, Messrs. W. Anderson and G. Kapp, Professors 

 J. Perry and R. H. Smith, with the present writer as Secretary, who, at 

 the meeting in the following year, was requested by the Sectional Com- 

 mittee to draw up and complete the report. Owing to causes which 

 were explained in a letter to the Committee last year, he was unable to 

 do this in time for the following meeting, and now begs to present the 

 result of his work up to the present time on the subject. 



The preliminary report gave a list of a number of works which 

 avowedly dealt with graphic methods, but at the same time it was stated 

 that a large number of standard works based to a great extent on such 

 methods, or in which such methods were frequently applied, were 

 excluded from the list. 



It does not require much acquaintance with the subject to realise that 

 there is scarcely a treatise or publication dealing with mechanical science 

 that does not employ some kind of graphical expression as a means of 

 exposition or calculation. How great the range is may be gathered from 

 a consideration that graphic methods really include every way of repre- 

 senting numerical quantities by means of drawing, except 1. The repre- 

 sentation of actual bodies ; and 2. Purely geometrical construction. 

 Hence we have to go back very early in the history of mechanical science 

 to find the first graphic treatment of the subject. 



A study of the list of references to graphic applications in the scientific 

 literature of this country appended to this report will show that gradu- 

 ally, but with an ever-increasing rate, it has become the custom to employ 

 representations of numerical results graphically, and to suggest and use 

 the solutions of various problems by graphic methods. 



These two purposes, viz. — 



1. Representation of results by means of plotted curves and auto- 



graphic diagrams ; 



2. Solution of problems by graphical methods, 



seem to suggest a satisfactory way of dealing with the subject. At the 

 same time, no hard-and-fast distinction must be insisted on, because a 

 curve, although chiefly useful for the purpose of representing the general 

 results of an experiment or calculation, may be employed, by means of 

 interpolation or otherwise, to effect a calculation. 



(1) It would be quite legitimate in this report to trace the develop- 

 ment of the represention of results, but the best method of procedure is 

 to state the results which have been achieved, as this aspect is the 

 practical and useful one. 



Inasmuch as every possible kind of numerical quantity can be plotted, 

 it would not be possible, even if it were desirable, to give anything like a 

 complete list of the subjects which have been, or may be, plotted. 

 Further, in dealing with instruments used for the purpose, there must be 



