24 Notes on the Design of Telescope Tubes. 



tension of the upper side and a longitudinal compression of 

 the lower side. To resist these stresses we have a series of 

 curved bars placed at an angle of about 30° with the lines 

 of stress. These on the upper side tend to straighten when 

 under stress, and those below become more curved. Hence 

 arises a general bulging in of the upper or extended side, and 

 a general bulging out of the compressed or lower side of the 

 tube. This action is plainly visible in the model when 

 loaded. Those parts of the tube which connect the top 

 and bottom together jtre subject to equal inclined stresses — 

 the bars that slope upward toward the open end to com- 

 pression, the others to tension ; the former tend to become 

 more bowed, the latter to straighten; and as they are 

 riveted at each intersection^, these two actions probably 

 antagonise and balance each other. 



The angle-iron rings which are placed at intervals along 

 the tube do not, as far as I can see, fulfil any important 

 function. I think the tube would be improved much if 

 they were removed, and longitudinal booms inserted 

 instead. 



In order to verify experimentally the preceding conclu- 

 sions, the two cardboard models represented by Figs. 1 and 

 2 were constructed. They are of equal length, and will 

 permit the unobstructed passage of cylinders of rays of light 

 of equal diameter. They were constructed from the same 

 sheet of cardboard, special care being employed to use an 

 exactly equal area of cardboard in each model ; and both in 

 constructing and testing them every possible precaution was 

 taken to place them under absolutely identical conditions. 

 The test load was a weight of 12 ounces avoirdupois, applied 

 at right-angles to the length of the tube at its upper or free 

 end, the other end being firmly fixed to a massive frame. 

 After each experiment the tube was rotated on its axis, so 

 that the test load should act on a difterent plane. In this 

 way Fig. 1 was tested six times with the test load acting in 

 planes passing through two opposite angles, and six times in 

 planes passing through the centre of two opposite sides ; and 

 Fig. 2 eight times in various directions equally distributed 

 round the circle. The mean results of these experiments 

 were as below : — 



Fig. 1. Deflection over angles, mean of 6 results, "0825 in. 

 „ „ „ sides „ „ '0314 in» 



Fig. 2. „ mean of 8 results, "0876 in. 



