448 THE STRENGTH 



same way, but the results were so anomalous 

 that no general conclusion could be drawn 

 from them. There was, however, one fact which 

 so often presented itself, that it could not I 

 conceived arise from accident — the piece fre- 

 quently broke at a distance from the shoulder, 

 shewing clearly the greater weakness in that 

 part where the flanges approached to equality. 

 This I noticed the more particularly, as, 

 according to the reasonings of Mr. Tredgold, 

 the strongest form of cross section was that 

 where the top and bottom ribs or flanges were 

 equal. 



36. The indecisive character of these experi- 

 ments, Mr. Ewart was of opinion, might arise 

 from their having been made on too small a 

 scale, and he suggested to me the propriety of 

 repeating them on a larger one: which was 

 done, and the results shew a degree of uniformity, 

 which forms a striking contrast to the others. 

 The mode however of making the experiments 

 was varied ; for, as I had met with difficulty 

 in getting sufficiently good castings from the 

 form of model before used; and as Messrs. 

 Fairbairn and Lillie had a very convenient 

 apparatus, a long lever, for trying or breaking 

 beams, I felt rather more desirous of making 



