628 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



Description of Apparatus. 



Each of the iron plates mentioned above was brazed at each end 

 into a thick copper block, as in Figures 7 and 8. After being so 

 fixed each plate- was measured in thickness at nine points, the distri- 

 bution of which is well shown by Figure 5, the distance between (1) 

 and (2), (2) and (3), (4) and (5), etc., being appoximately 1.5 cm. 



Figure 5. Figure 6. 



Plate 1. — The measurements of Plate 1 were 



Length between coppers .... 5.1cm., 

 Width " " .... 1.98 cm., 



Arms about 0.12 cm. long and about 0.10 cm. wide. 



cm. Ft. 



r0.1490 at 1 

 Thickness <^ 0.1490 " 4 



1 0-1475 " 7 

 Means 0.1485 



cm. Pt. 



0.1460 at 2 



0.1480 " 5 



0.1450 " 8 



cm. Pt. 



0.1505 at 3 



0.1495 " 6 



0.1500 " 9 



0.1463 



0.1500 



As the part of the plate which is particularly important in the 

 transverse effects, the effects that interest us most, is the narrow 

 transverse strip including the arms, the thickness of this part is to be 

 especially considered. This, according to our measurements, is about 

 0.1463 cm. ; but as the method of measurement, by means of vernier 

 calipers with faces of considerable width, must tend to give a maximum 

 value, we take off about 1 per cent from the quantity just given and 

 call the mean thickness of the median transverse strip 0.145 cm. 



To Plate 1, eight wires were attached (Figure 6) by means of copper 

 electrolytically deposited. Four of these, Ci, C2, C3, C4, were of copper 

 about 0.017 cm. in diameter, the others, Ii, I2, I3, I4, were of annealed 

 iron wire about 0.020 cm. in diameter. This iron wire was drawn from 

 a portion of the same bar that furnished the plates 1 and 2. These fine 

 iron wires were joined to somewhat heavier iron wires of the same source 

 15 cm. or more above the iron plate, and the heavier wires, perhaps a 



