IRON, STEEL, AND NICKEL TUBES IN THE MAGNETIC FIELD. 533 



tube, it would be ill-spent labour to attempt in any way an evaluation of the mean 

 magnetisation or induction along the tube. The volume changes that are here being 

 studied depend, no doubt, upon the magnetisation rather than upon the magnetising 

 force. And it is not improbable that the complex distribution of magnetisation has a 

 great deal to do with many of the peculiarities of strain which declare themselves 

 and which are discussed below. These considerations convinced me that any attempt 

 to institute a comparison between volume changes and a purely hypothetical estimate 

 of magnetisation could lead to no satisfactory issue. Accordingly in what follows 

 the comparison is, in general, made between volume change and the magnetic field 

 at the centre of the six-fold coil. 



§ 7. Method of Experiment. — As already noted, the tube stood vertical in the 

 heart of the magnetising coil. The capillary rose vertically out of the cap high enough 

 to clear the top of the coil, and was then bent into a horizontal position, all except the 

 small part near the end, which was turned down (see § 5). The liquid meniscus was 

 formed in the way above described (§ 5) at a convenient part, and sighted by the 

 microscope. A strong light reflected from a mirror placed beneath rendered the 

 meniscus clearly visible. Its position could be estimated to tenths of the small 

 divisions of the micrometer scale, eighty of which nearly filled the field of view. 



The magnetic field was then established, and the resulting displacement of the 

 meniscus noted. On removal of the field, the meniscus moved back towards its original 

 position. When the magnetising force was applied and removed for the first time, the 

 meniscus did not move back to its original position, indicating a permanent change of 

 volume. The same field was then applied in the opposite direction and removed, the 

 corresponding meniscus displacements being noted at every operation. After a few 

 successive applications of this cycle of alternating positive and negative fields, the tube 

 arrived at a steady cyclic condition in regard to the changes of volume. When this 

 cyclic state was reached, the displacements of the meniscus were almost exactly the same 

 both at make and break of the magnetising current. Also, unless some magnetic bias 

 existed in the tube — the after-effect of a previous magnetisation — the amount and 

 direction of displacement were independent of the direction of the magnetic field. # The 

 mean of the four displacements at make and break of the direct and reversed currents is 

 taken as the change of volume due to the corresponding field cyclically applied. It was 

 difficult to get rid altogether of magnetic bias in the lower fields ; but by beginning 

 with the highest fields and coming down by small enough intervals to the lower 

 fields, we managed to keep the tube in a fairly neutral condition. On working up 

 again to the highest fields we got practically the same volume change for any given 

 field as was got originally in the descending series. The magnetising current was 

 measured on a Kelvin Long Grade Galvanometer, which was calibrated from time to 

 time in situ by direct comparison with a Kelvin Ampere Balance. 



Simultaneously with the observations of volume change, measurements were taken 



* The earth's vertical field is negligible in comparison with the strong fields necessary to produce the strains hej e 

 discussed. 



