224 



ANNUAL REPOET SMITHSONIAN INSTITUTION, 1912. 



and of Perkins. Suffice it to say that the results vary between 70 

 and 235 for the atomic weight. Debierne, employing Bunsen's method, 

 which is based on the escape of gas through a narrov/ orifice, gives 

 the figure as 220. Now, taking into account the fact that radium 

 in changmg to niton gives off an atom of helium, and accepting for 

 the atomic weight of radium 226.4, found by Madame Curie and con- 

 firmed by Thorpe, the atomic weight of the emanation must be 

 222.4 = 226.4-4. 



The molecular weight is fixed by the density, and in the case of a 

 monatomic gas (and there is every reason to believe that niton is 

 monatomic) the molecular weight becomes identical with the atomic 

 weight. As a definite test it was necessary to weigh a known volume 

 of niton. 



But how is it possible to weigh a gas of which the largest available 

 quantity did not exceed one-tenth of a cubic millimeter? I have 

 already hinted that, with the aid of the microbalance invented by 



Steele and constructed by Whytlaw- 

 Gray, the idea is not chimerical. Let 

 me give you a sketch of the methods 

 we have employed. 



First, however, I must make clear 

 the construction of the balance. Sci- 

 entists have for some years been 

 employing a new substance — melted 

 quartz. Everyone knows how re- 

 sistant it is. Its coefficient of ex- 

 pansion by heat is almost zero, and 

 we may work it like glass, drawing out rods of suitable thickness. 

 Before constructmg the balance, I consulted my colleague, the pro- 

 fessor of engineering, as to the best form for a bridge to permit it to 

 resist a maximum pressure. He was so kind as to give me a draw- 

 ing of it. 



To construct the balance-beam grooves are drawn upon a plate of 

 graphite with a knitting needle, and in these short rods of silica of 

 about a half millimeter in diameter are placed. Wliere the ends of 

 the rods touch, they are melted by turning on them for a moment 

 the flame of an oxyhydrogen blowpipe. The knife-edge is formed 

 of a small drop of silica, melted at the end of a short rod and ground 

 to the form of a wedge with the greatest care. Seen under a micro- 

 scope it must be straight, without notches, and well polished. Per- 

 pendicular to this rod and quite near the knife-edge is sealed a second 

 rod which serves to adjust the knife-edge so that when the rod is 

 sealed to the beam of the balance, the knife-edge forms a right angle 

 with the plane of the beam. This second rod serves also to support 

 a small mirror of platinized silica reflecting the light of a Nernst 



Fig. 2. 



