=61.= A very useful electrometer of great sensibility has been devised by Dolezalek[106]. It is of the ordinary quadrant type with a very light needle of silvered paper, spindle shaped, which lies fairly close to the quadrants. A very fine quartz suspension is employed. In consequence of the lightness of the needle and its nearness to the quadrants, it acts as its own damper. This is a great advantage, for difficulties always arise when the wire dips into sulphuric acid, on account of the thin film which collects after some time on the surface of the acid. This film obstructs the motion of the platinum wire dipping into the acid, and has to be removed at regular intervals. These instruments can readily be made to give a sensibility of several thousand divisions for a volt when the needle is charged to about one hundred volts. The sensibility of the electrometer passes through a maximum as the potential of the needle is increased. It is always advisable to charge the needle to about the value of this critical potential. The capacity of the electrometer is in general high (about 50 electrostatic units) but the increased sensibility more than compensates for this. The needle may either be charged by lightly touching it with one terminal of a battery, or it may be kept charged to a constant potential through the quartz suspension. Dolezalek states that the fibre can be made sufficiently conducting for the purpose by dipping it into a dilute solution of calcium chloride or phosphoric acid. I have not found this method satisfactory in dry climates as in many cases the fibre practically loses its conductivity after a few days exposure to dry air. In addition to its great sensibility, the advantage of this instrument is in the steadiness of the zero and in the self-damping. A sensibility of 10,000 millimetre divisions per volt can be readily obtained with this electrometer, if a very fine fibre be used. The use of such high sensibilities cannot, however, be recommended except for very special experiments. The period of swing of the needle under these conditions is several minutes and the natural leak of the testing vessels employed, as well as electrostatic and other disturbances, make themselves only too manifest. If measurements of minute currents are required, an electroscope of the type described in Section 56 is much to be preferred to a very sensitive electrometer. The electroscope readings in such a case are more accurate than similar measurements made by an electrometer. For most measurements in radio-activity, an electrometer which has a sensibility of 100 divisions per volt is very suitable, and no advantage is gained by using an electrometer of greater sensibility. If still smaller effects require to be measured, the sensibility may be increased to several thousand divisions per volt.