436 Remarks on the Article " Strength of Materials " &fc. 



curs in Dr. Robison's account of them. There were only two 

 bars cut one-third through, and only two cut half through ; see 

 Du Hamel's Transport du Bois, p. 421. Paris, 1767. 



The author prepares us for his own experiments and calcula- 

 tions by observing, " that Galileo is much nearer, if not exactly 

 conformable to the actual operation." But without giving area- 

 son for this opinion, he gives a rule founded on experiments 

 which are not detailed, and it is probable the rule will require 

 to be varied to suit different kinds of materials, though the only 

 variables that ought to have place in it, viz. the direct cohesion 

 and extensibility, are both included. The autlior must be aware 

 that a rule of this kind, however correctly it may agree with 

 one set of experiments, will not agree with another; therefore, 

 such a rule must be empirical. Besides, if it he necessary to 

 resort to a rule of this kind, it ought to be simple; which cannot 

 be said of the present one. Several rules have been given at 

 different times from the experiments of Buffon, and a very easy 

 one is given in Brewster's Encyclopaedia, art. Carpentry, p. 508; 

 but were these rules correct for timber nevvlv cut, they would be 

 incorrect for dry timber. To attempt to include accidental 

 qualities in a general rule is absurd, yet this must always be the 

 case with rules founded on experiment. 



The author remarks, that the deflexion " is not a necessary 

 datum in estimating the strength of timber for any practical 

 purposes of building." Had he said, it is the only one that is really 

 useful, he would have been much nearer the truth. See Dr. 

 Young's Nat. Philosophy, vol. i. p. l.'jS; or his paper on Sep- 

 ping's Method of Ship-building, Phil. Trans.; or Buchanan's 

 Essay on the Shafts of Mills, p. 73. 



If we had a correct theory, as far as the direct cohesion and 

 extensibility of the material are concerned, and four or five ex- 

 periments, under different circumstances, on each kind of ma- 

 terial, the material being accurately described, its specific gravity 

 and extensibility correctly ascertained, and the manner of making 

 the experiment accurately detailed, — then we should have a series 

 of results under various circumstances, which would enable th« 

 mechanic to apply the theory to practice, with such limitations 

 &s a comparison of his material with that of the experiment 

 would readily point out to him. 



As long as the practice of giving mean results without de- 

 tailing the experiments shall continue, so long may we expect 

 to remain ignorant of the nature of the resistance of solids. In- 

 stead of following the methods of Prony or Laplace, the example 

 of the chemical writers ought to be followed. This will lead us 

 insensibly to perfection, while that is but a specious covering for 

 ignorance. ' I do 



