Ill 



Tile above tisiire is t'iiso 111. and ^^howis the i:r(auid a in)rabola in longi- 

 tudinal section. This case most nearly coincides with the actual condi- 

 tions, for haul, of any of the three cases, and any variance discovered here 

 in the results of the four nietliods are about what would occur in actual 

 praitice, while those discovered luider Case II are limiting values. 



It being the object of the writer to discover the greatest variance that 

 could occur, thus obtaining limiting values, most of the computations were 

 under Case II, with a few tesl investigations under Case III, to obtain 

 values that would be encountered more often in practice. 



Method Iso. 1 depends u])on the general form that the centi-r of grav- 

 ity of individual prismoids is located a distance from llu' inid-s,'ition to- 

 ward the larger section a distance 



^ r, A — A' 

 For all practical purposes it is exact. 



1 A — A' :i;Monieuts 



and Haul =r 



-Volumes. 



