Chemistry and Physics. 457 



opening was situated on the optic axis of a double-convex quartz 

 lens and at such a distance from the lens as to cause the ordinary- 

 light waves and shorter infra-red radiations — corresponding to 

 refractive indices between T43 and 1'55 — to form divergent 

 beams after emerging from the lens. On the contrary, the very 

 long heat waves were brought to a relatively sharp focus at a 

 similar aperture in a second screen. By stopping out the central 

 zone of the lens by means of a black paper disc, only the longest 

 waves were allowed to enter the hole in the second screen. A 

 second lens, similarly stopped and placed behind the last screen, 

 completed the purification of the radiation and also brought the 

 waves to a focus on the thermocouple of a radio-micrometer. 

 The wave-lengths and distribution of energy in the isolated radia- 

 tion were determined by means of a quartz interferometer placed 

 close behind the second aperture. 



The mean wave-length was found to be about 108/a and the 

 existence of waves as long as 150//, was established. The longest 

 waves previously recorded had a length of 97/u. 



Having determined the mean wave-length of the isolated radia- 

 tion, the authors studied the absorbing and reflecting powers of 

 various solids, liquids and vapors for this region of the spectrum. 

 Only a few typical results may be mentioned here. Diamond 

 was remarkably transparent, while fused quartz and water-glass 

 were very opaque. Black paper and even black cardboard were 

 partially transparent. A thin lamina of mica densely smoked 

 was exceedingly transparent to waves of length 108/x. As pointed 

 out in the paper, this fact is of importance in the construction of 

 radio-micrometers, etc. The reflecting power of water was found 

 to be comparatively high, which is probably due to the presence 

 of one or more absorption bands in the region under investiga- 

 tion. — Phil. Mag. (6), xxi, 249. h. s. u. 



6. A Jfethod of Calibrating Fine Capillary Tabes. — The 

 usual method for the determination of the mean area of cross- 

 section of the bore of a capillary tube decreases in precision as 

 the radius becomes smaller because of the difficultj^ of weighing 

 the thread of mercury employed to a sufficiently high degree of 

 accuracy. For example, the mass of a column of mercury 10 cm 

 long contained in a capillary tube of - l mm internal diameter is 

 about - 01 gram, so that to obtain an accuracy of 1 per cent in 

 the square of the radius the weighing must be correct to - l mg. 



A very promising method, which depends primarily upon the 

 determination of the electrical resistance of a thread of mercury, 

 has been devised and tested by T R. Merton. Each end of the 

 capillary tube is fitted into a rubber stopper, and each stopper 

 enters the shorter arm of a glass tube shaped like the letter L 

 and of relatively large internal diameter. The longer arms of 

 the L-tubes are vertical when the capillary is horizontal. The 

 entire capillary bore and the greater part of the auxiliary end 

 vessels are filled with pure mercui-y. By using a suitable air 

 pump for evacuation, no difficulty is experienced in forcing the 



