CoLLKt'TlO.NS IN r.CONOMIO (JKOLOCJY AND M i: I Ai.i,ri;(; V. Ifiy 



ami till' small pirct's, with tli<»s»' ol' a ImiMlrt'd iiiflu-s aln-iidy ^i\«Mi 

 iimlrr si'iics I). 



Di'scriptioti. 



I'outractioD or iireii. 



Lnrce 



Small 



J... 



Loug 



Mean. 



ritiniatAext«iiRtnti. 



Large 



Nprtps 



Small 



(scrlrs 



Tliese results show tla^ hirge specimens, which liehl the highest stress, 

 have contracted h'ast, and the lon^i" specimens, which hcdd tlie lowest 

 stress, have contracted most, thus to some exteat accounting for the 

 difterences in the stress. 



Series F. 



To ascertain the mechanical effects produced on steel plates of various 

 thicknesses by holes, some drilled, others luinched, tests under pull- 

 ing stress were made. The same mateiial as scries I), K, and F was 

 trst«'<l. 



There are tliree rows of live each <»t' rivet holes, A inches t)etweeii 

 their centers, the pitch of the five holes across the plate being 2A 

 inches. 



In the fiv«' unannealed plates with drilled holes the loss ol U'lisile 

 strength ranges from 21.72 to 25.1.'», the average being 23.21 per cent.; 

 and those annealed from 23.04 to 27. OS, average 24.'.»2 jicr <rent., yield- 

 ing a total average of 24.00 per cent. The loss in no instance was 

 nearly so great as that representing the material removed, rln- mean 

 being 24. Of. against .'.O.SO, or (».74 per cent. less. 



In the live unannealed plates with puiM^hed holes the loss of tensile 

 strength ranges from 30.50 to 40.04, the mean being .'57..S.~». The greatest 

 loss, 40.04, occurs in the ^-inch plate, and is 4."). 17 in I he jincli,and least 

 in the ^fjinch jdate, whi(!h the other experiments proved to be the softest 

 of the Ave, and con.sequently the punching had not acted soinjuriously 

 as in the other two. The loss from punching is not constant, but varies 

 with the thickness and also wirh the hanlness of the mat«M'ial. That 



