mum exciting field. After this 22 simultaneous determinations of H and B were made in each half-cycle, and both cycles were carefully plotted on such a scale that each was about 40 centimeters long. It appeared that the maximum values of the induction were almost identical, and that at no point were the plotted curves so much as 1 millimeter apart. In filling the same glass tube a number of times in succession from the same lot of filings, it was of course impossible to pack exactly the same mass into the same space twice; but the hysteresis diagrams,

many of which were obtained by Mr. J. Coulson and myself, were always of the same shape, and the intensity of magnetization due to a given exciting field seemed to be strictly proportional to the mass. The density of the untreated filings was only about four tenths of that of massive cast iron.

The demagnetizing effect of the ends in a rod of solid iron only fifty diameters long would of course be very serious, but in the present case, due to the relatively small value of I for a given value of H, it is far less important. In order to determine from my observations the permeability at the centre of an endless column of material like

mine, I have used the value of N computed by Du Bois 2 from the experimental results of Ewing, Tanakadaté, and himself.

Figure 2 shows half of a hysteresis diagram obtained by shearing very slightly a diagram obtained by Mr. J. Coulson and me from a column of untreated particles 50 centimeters long and 1 centimeter in diameter, which weighed just over 100 grams. This diagram has all the characteristics of the many others for which I have the materials.

As the figure shows, B was almost exactly 2100 for an exciting field of 255 gausses, and this corresponds to a value for I of only about 147. When the external field was removed, the remanent value of I was about 20.8; the coercive force, however, was comparatively large, being about 16. It is evident that the exciting field would need to be very strong to magnetize the column of particles approximately to saturation. The subject of the saturization of masses of iron filings has been discussed at length by Maurain and by Trenkle.

THE JEFFERSON LABORATORY,

HARVARD COLLEGE.

* Ewing, The Philosophical Transactions; Tanakadaté, The Philosophical Magazine, 1888; Du Bois, Annalen der Physik., 1892; "The Magnetic Circuit," chapter vi. For a criticism on this process, see Trenkle, 1. c.

Proceedings of the American Academy of Arts and Sciences.

VOL. XLII. No. 4. -JUNE, 1906.

CONTRIBUTIONS FROM THE JEFFERSON PHYSICAL LABORATORY, HARVARD COLLEGE.

ON THE LENGTH OF THE TIME OF CONTACT IN THE CASE OF A QUICK TAP ON A TELEGRAPH KEY.

BY B. OSGOOD PEIRCE.