May 8, 1903.] 



SCIENCE. 



733 



whether the contractions arise from the 

 nerves, nerve ends, the muscle substance 

 or a disturbance of the electrical equilib- 

 rium within the muscle mass. It must not 

 be forgotten also that the relaxation of the 

 muscle is possibly an active process, and, 

 as experiments indicate on cilia, the relax- 

 ation may be stimulated by the positive 

 ions, the contraction by the negative, some- 

 what as Howell suggested for potassium 

 and calcium in the case of muscle. Lingle 

 has already worked in this direction and 

 will no doubt be able to clear up some of 

 the discrepancies. 



10. I believe the results so far obtained 

 support strongly the truth of the hypoth- 

 esis of the antagonistic action of the anions 

 and cathions on protoplasm. They sup- 

 port Loeb's original suggestion of the im- 

 portance of valence and of my conclusion 

 that in motor nerves and some other tis- 

 sues the anion stimulates, while the cathion 

 inhibits. They also support the explana- 

 tion of chemical and electrical stimulation 

 and electrotonus given in my former paper. 



11. The results obtained indicate also 

 the truth of the general law, i. e., the 

 physiological aetion of any salt is equal to 

 the algeiraic sum of the actions of its ions. 



A. P. Mathews. 

 Chicago, 



March 3, 1903. 



8TEEMMAT0GRAPH TESTS. PRINCIFLEB 

 AND FACTS RELATING TO TEE DISTRI- 

 BUTION OF THE STRAINS IN TEE 

 BASE OF RAILS UNDER MOVING 

 TRAINS. 



Before it was possible to make any tests 

 of precision showing the distribution of 

 the stresses in rails under moving loco- 

 motives, it was necessary to improve the 

 tracks, and introduce stiffer rails than were 

 in use prior to 1884. The 4|-inch 65- 

 pound rails, with their splice bars, were 

 too weak to distribute the loads of the loco- 

 motives and cars in an efficient manner. 



While the distribution theoretically fol- 

 lows the same general law in the lighter 

 rails, yet the efficiency is so much less that 

 it is impossible to obtain comparative tests 

 in practice to confirm the theory of the 

 distribution of stresses under locomotives. 



On the 4^-inch rails, the heaviest axle 

 loads on passenger locomotives prior to 

 1889 were about 27,500 pounds. When 

 many miles of the stiffer 5-inch 80-pound 

 rails were in the track, the axle loads were 

 increased to 40,000 pounds per pair of 

 driving wheels. 



The stiffer rails permitted better joints, 

 capable of holding up the ends of the rails, 

 which continued the functional action of 

 a rail as a continuous girder to adjacent 

 rails. 



After the stiffer rails have been well 

 surfaced in the track, the portion under 

 the driving wheels becomes practically a 

 restrained beam with numerous supports, 

 the front end being held down by the for- 

 ward truck wheels, and the other portion 

 of the rail by the tender wheels. 



The stiffer rails have been laid upon the 

 same road-beds, without increase of width, 

 to distribute laterally the heavier wheel 

 loads to more breadth of road-bed. 



The track for steam railroads is by con- 

 struction flexible, but notwithstanding the 

 high standards of smoothness which have 

 been secured by reducing the looseness of 

 the superstructure and its flexibility to 

 small limits by stiffer rails, it is not a 

 limited flexible structure like a bridge, in 

 which the strains in the members may be 

 analyzed and calculated. 



The problem— or series of problems — in 

 reference to the strains in rails and their 

 distribution under moving locomotives and 

 cars, is so complicated by the looseness of 

 the superstructure apid the imperfect elas- 

 ticity of the road-bed, that it has not 

 yielded to mathematical analysis, as for 



