1895.] in the Magnetic Permeability of Iron. 229 



One objection to this explanation was, that as the iron bolts 

 holding the plates together were heated nearly as much as the plates,, 

 and therefore expanded nearly as much, the pressure could only vary 

 slightly. 



It was thought that in some forms such a gradual hardening 

 might also be caused by the pressure exerted by the weight of the 

 upper plates on the lower ones. But the increase was found where 

 there was no superincumbent weight, in transformers in which the 

 plates rested on their sides or edges. 



In order to investigate this point two experiments were tried. The 

 first was with the sample marked No. 5. In order, as far as possible,, 

 to avoid any compression, the plates in this sample were not bolted 

 or clamped together, but, as described above, were merely held 

 together by a binding of insulating tape and by the layer of copper 

 wire with which they were covered. The results given in Table V 

 do not differ materially from those of Nos. 1, 2, and 3. 



An observation which tends to indicate that pressure is not the- 

 cause of the change of condition was made some time ago. Different 

 parts of the iron of a transformer, which had been in use some time, 

 were examined, when it was found that the middle portions were 

 rather more affected than the end portions. As the end portions,, 

 being most exposed, are least heated, and as in this instance all parts 

 were subject to the same amount of pressure by the clamps or bolts, 

 this observation does not support the supposition that pressure is the 

 cause of the change. As, however, commercial iron varies in ibs 

 hysteresis, even in the same sheet, the observation is of doubtful 

 value 



A second set of tests is that marked No. 6. In this case the iron 

 used was of the same size as in Nos. 1, 2, and 3, but of smaller 

 amount. 



It was arranged with a testing coil in the same manner, but instead 

 of keeping the iron with its coil complete throughout the experiment,, 

 the coil was only put on for the tests, and was then removed, the 

 iron being placed in the oven, arranged as a flat block (its upper 

 surface being about 150 sq. ins.), with a weight of about 300 Ibs. on 

 it till the next test was made. It was thought that perhaps the 

 slight pressure produced in this way might affect the results. These 

 are given in Table VI and Diagram No. 6, and are markedly different 

 from the tests of Nos. 1, 2, 3, and 5. The increase of hysteresis is, 

 greater, although the test has not been so long in progress. 



This point deserves full investigation, and the author hopes to 

 obtain further results from tests now in progress.* 



Change of Power-factor. The power-factor or ratio of watts to 

 volts x amperes shows an irregular increase in all the tests. This 

 * See Appendix I, p. 240. 



