they were free from magnetism. The needles were placed in various directions in the plane of the magnetic meridian, sometimes in the angle of the dip, sometimes perpendicular to the magnetic meridian, and also at various angles with regard to it. In some cases the heads of the needles were exposed in place of the points, to the violet rays. Perhaps it might have been expected that the influence would have been greater in those instances in which the needles were placed in the plane of the magnetic meridian, and at the angle of the dip; and, consequently, polarity might have been expected to take place in a shorter time under these circumstances; yet in fact there seemed to be no difference; most of the needles became magnetic, some in longer, others in shorter periods, varying from about half an hour to four hours, but depending on circumstances which I have not yet been able to detect, further than that a number of results induced me to believe, that the experiments were more successful from ten to twelve, or one o'clock, than later in the day. The portion of the needle exposed was almost always a north pole, whether it pointed upwards or downwards. In a few instances in which the contrary occurred, it may possibly have arisen from some previous disposition in the needle to magnetism, too slight to be observed.
The distance of the needle from the prism was frequently varied by fixing the needle to the wooden pole of a firescreen, but without material variation in the effect. I found it unnecessary to darken the room; it was sufficient to place the prism so as throw the spectrum on any place out of the sun's rays.
My next object was to endeavour to ascertain whether any
