92 W. H. WARREN. 
as small as one two hundred and fifty thousandth of an inch, It 
has the advantage also of not being influenced by the tempera- 
ture of the body of the observer to anything approaching the same 
extent as with micrometer readings, and is probably the most 
accurate apparatus yet designed for measuring the small defor- 
mations which occur within the elastic limit of materials. 
The apparatus is illustrated in Figs. 2, 3, and 4,’ which show its 
application to the testing of a mild steel bar, and a cube of concrete. 
The following table gives a series of readings taken with the 
apparatus for a round specimen of mild steel :— 
aeeien) & | Met 98) bl 
a: bas| A | |g 2 | Remarks. 
aes ne 3 Left Right g 2 | = 
ASA) 8st) A lap m™m),5mm| af | a 
20 |200| 025! 0, 0,1 O| 0 
100) 80) 88 | 168 |168 
2:00 | 120; 200} 306 | 228 
3:00 | 299 | 323 | 622 | 266 
4:00 411 441 852 | 230 
500 | 530 | 551 | 1081 | 229 
6-0 64 660 | 1309 | 228 
7:00 | 758 | 779 | 1537 | 228 
8-00 | 872 | 894 | 1766 | 229 ee 
—L imitof elasticity. 
9:00 | 995 | 1009 | 2004 | 238 
10-00 | 1140 | 1130 | 2270 | 266 
10°25 | —Yield point 
16°50 |—Breaking Load. 
Mr. Béhme has also made a pair of Roller Extensometers for 
the Engineering Laboratory, which will be used for ascertaining 
the deflections of beams, and compression of columns. This 
extensometer consists of a dial divided into five hundred parts and 
a rotating index, which has an angular displacement proportional 
to the deformation of the test piece. One revolution corresponds 
with one centimeter deformation, so that readings are taken to 
(#5 mm.) one-fiftieth part of a millimeter, or one thousand two 
1 The letters in these figures refer to Mr. G. H. Knibbs’ paper 00 the’ 
same subject, following on. : 
