

168 Proceedings, &c K , for 1884. 



In the discussion which followed, Professor Kernot mentioned 

 that, as a general rule, he was averse to spreading the metal in 

 the form of thin plates, as is done in web girders, on account of 

 corrosion, but in some cases their use is unavoidable. From a 

 series of experiments on cardboard models, he has come to the 

 conclusion that their behaviour is very different when loaded on 

 the top and on the bottom member ; but he is not aware of any 

 difference of construction in the two cases. Further discussion 

 was postponed till next meeting. 



Mr. Behrendt then exhibited a telemeter, a new instrument to 

 serve the purpose of a stadiometer, but on a base of only 18 

 inches in length. The inventor claims that the error in measuring 

 a length of 1000 yards is less than two feet. 



From a series of careful experiments, carried out under the 

 superintendence of Mr. A. 0. Allan, who has had considerable 

 experience in careful observation while carrying out the geodetic 

 survey of this colony, it appears that at so short a distance as 15 

 chains the instrument cannot be depended on to within 10 or 

 15 feet. 



At a distance of 1000 feet an error of one second in the 

 measurement of the angle, subtended by a distance of 18 inches, 

 is equivalent to an error in the apparent distance of the object, of 

 over three feet. So that very accurate results can hardly be 

 looked for. 



Professor Kernot showed some photographs of the Tay Bridge, 

 and some of the rolling stock, taken after the accident. Then 

 followed a few words about the bridge, the cause of the accident, 

 and a general discussion on the "Factors of Safety" to be used 

 for wind pressure. 



May 21s*, 1884. 



Mr. Behrendt opened the discussion on Mr. Fraser's paper. 



He pointed out concisely the action of a load, distributed in 

 any manner, on a beam ; giving the mathematical expression for 

 the various results. Though no English work contains any 

 examination of the web stresses, yet the subject is very ful]y 

 considered in many German works, a few of which Mr. Behrendt 

 mentioned. 



Professor Kernot pointed out the desirability of experimenting 

 on models ; these should be of iron rather than paper, since paper 

 has no lateral stiffness to resist the crumpling action. 



Several members joined in the discussion. 



