﻿448 Barus and Strouhal — Viscosity of Steel. 



with a single adjustment, the viscous movement can be traced 

 with extreme accuracy for an indefinite length of time. There 

 is no doubt that in some of the cases examined such motion 

 will demonstrably exist even for a year after the torsion has 

 been imparted. Kohlrausch's method is an absolute method ; 

 but his wire must be twisted and untwisted once for each 

 observation. The method of logarithmic decrement is diffi- 

 cult of interpretation, is really a complex method and limited 

 to short intervals of time. 



In experimenting with glass it was necessary to modify the 

 apparatus so that twists less than ^=180° could be applied. 

 This is easily done by perforating the screw of the upper bolt 

 and passing the glass fiber through the hole. Torsion in any 

 amount may then be applied to the projecting fiber. 



2. Our second (tubular) apparatus combines with the advan- 

 tages of the one just sketched, the ulterior desideratum of en- 

 abling the observer to follow the viscous detorsion immediately 

 after applying stress ; and therefore also of exhibiting the 

 effect of positive torsion immediately succeeding negative tor- 

 sion and vice versa; or of studying the effect produced by an 

 indefinite number of alternations of the sign of the twist. In 

 the second form of apparatus the wire is introduced into a nar- 

 row tube just large enough to surround it. The tube is fixed 

 above and fastened to the wire below. Twist is imparted to 

 the wire and through it to the tube. The observations, in 

 other respects, are conducted as above. 



Method of Observation. — In both forms of apparatus, angles 

 are registered by Grauss' method of telescope* and scale. We 

 secure extreme accuracy by using two mirrors, one of which 

 (stationary) furnishes a reliable fiducial mark for the other 

 (movable). These mirrors are easily so adjusted that their 

 respective scale-images appear simultaneously in the field of the 

 telescope one above the other and intersected by the same 

 cross-hair. Indeed it would not be difficult to arrange this 

 device that the stationary image could be used as a vernier on 

 the moving image. 



Method of Tempering. — The wires used (Stubs' steel) were 

 hardened galvanically as described elsewhere, f To Professor 

 IL A. Rowland, who placed his dynamo-electric machine at our 

 disposal for this purpose, and favored us with the benefit of 

 his advice during almost the entire afternoon, we owe our very 

 cordial thanks. 



The annealing was effected by drawing the glass-hard wires 

 through a zone of constant temperature by clockwork.^ 



* Three wood screws, the head of one of which is planed, of another hollowed 

 out conically. make a good adjustable plane-dot-slot arrangement, 

 f Bull. U. S. G. S., No. 14, p. 29 ; Wied. Ann., xi, p. 932. 1880. 

 % This Journal, xxxii, p. 2*79, 1886. 



