COMPLEX STRESS DISTRIBUTIONS IN ENGINEERING MATERIALS, 22] 
891 to 1-78 cm.; thickness from -041 to -135 cm. Length from 8 mm. to 
210 mm. Results seemed to indicate that when length>6 diameters, 
strength practically constant. Below that length, strength appeared to be 
inversely proportional to length, though it is stated that experiments are not 
sufficient to determine definitely the relation. 
6 Carman, A. P., 1906 Resistance of Tubes to Collapse. ‘ Univ. of Ill. Bull.,’ 
and Carr, M. L. vol. 3, No. 17, June 1906. ‘Sci. Abs.,’ 1906, No. 1986. 
Describes tests carried out on a number of lap-welded steel tubes, seam- 
less cold-drawn steel tubes and brass tubes. 
Following formule are given :— 
(a) When ; less than -025 
p— 25,150,000( 7) for brass tubes. 
p= 50,200,000( 5) for seamless steel tubes. 
(6) When ; > 0-03 
p = 93,365 ; — 2,474 for brass tubes. 
p = 95,520 ; — 2,090 for seamless cold-drawn steel. 
p = 83,270 ; — 1,025 for lap-welded steel. 
The last formula in each series agrees well with those obtained by Stewart. 
7 Clark, D. K. Strength of Boiler Flues. ‘ Engineering,’ vol. 46, p. 280. 
From reports of Manchester Steam Users’ Assoc. of six boiler-flues actually 
collapsed, gives formula 
50,000 
p=t( ; — 500) 
8 Fairbairn, W. 1858 Resistance of Tubes to Collapse. ‘ Phil. Trans.,’ 1858, 
p. 389. ‘ Brit. Assoc. Report,’ 1857, p. 215. 
Tests carried out on 32 wrought-iron tubes, lap-riveted, varying in 
diameter from 4 to 12 inches and in length from 1 foot 3 inches to 5 feet. 
Uniform thickness except in five cases, -043 inch. Results of tests repre- 
sented by formula 
= 9,675,600" 
Real vane did 
(p in lbs. square inch, ¢, 7, d in inches). 
9 Grashof, F. 1859 W. Fairbairn’s ‘ Versuche tiber den Widerstand von 
Rohren gegen Zusammendruckung.’ ‘Zeitschr. des 
Vereines deutscher Ingenieure,’ 1859, p. 234. Tod- 
hunter and Pearson, ‘ History of Elasticity,’ vol. ii. 
p.1; i, p. 666. 
From Fairbairn’s tests, deduces formule 
p08 a 
(a) p=1,033,620 posi qe fF thin tubes. 
(b) p=24,481,000 © for thick tubes. 
Ldtma 
Also considers a tube of slightly elliptical cross-section, and, using a 
method previously suggested by Bresse, obtains the formula 
