A.— MATHEMATICAL AND PHYSICAL SCIENCES. 43 



to feel much satisfaction with the special ad hoc hypotheses which have 

 had to be introduced into it. 



A stream of particles with a charge of one sign only is open to the 

 criticism, first put forward by Schuster, that the stream will dissipate 

 itself by electrostatic repulsion, and loses the hard outline which is one of 

 the most essential features. Lindemann has proposed to get over the 

 difficulty by making the stream neutral on the whole, still consisting, 

 however, of charged particles of both signs. Here, however, we lose too 

 much of the magnetic flexibility of the stream. Chapman proposes to 

 retain a slight excess of charge of one sign, and in this way is able to arrive 

 at a tolerable compromise. But one feels that more experimental 

 guidance is badly needed before we can venture with confidence into these 

 theoretically dark regions. The search for direct evidence might not seem 

 at first sight very hopeful, but not long ago a sensational suggestion was 

 made bj^ Stormer. His attention was drawn by Hals to echoes heard 

 after short-wave (131 metres) wireless signals sent out from Eindhoven in 

 Holland. These echoes have been found by Stormer and Hals at long 

 intervals up to as much as fifteen seconds after the original reception. 



Now, if we bear in mind that with the velocity of light the longest 

 terrestrial distances only give intervals about 1 of a second, it seems 

 inevitable that some extra-terrestrial reflector should be looked for. 

 Stormer finds this in the corpuscular stream as bent round by the earth's 

 magnetic force. Though the boldness of the idea is staggering, it is difficult 

 to suggest any alternative view. Stormer states that ' the variability of 

 the phenomenon indicated by the observations agrees well with the 

 corresponding variability of aurora and the magnetic registrations.' 



T. L. Eckersley has made an observation on electrical disturbances of 

 natural origin which he interprets as analogous to Stormer's. A 

 click is heard in a telephone attached to a large aerial, which is followed 

 at an interval of about three seconds by a ' whistler ' or musical note of 

 short duration. Further whistlers follow at intervals of 3-8 seconds, 

 each more drawn out than the previous. The musical notes are regarded 

 as due to the spreading action of a dispersive medium on an electrical 

 impulse. It is only at times of magnetic storm that these phenomena are 

 frequent. 



Further development of observations of this kind will be awaited with 

 keen interest. 



To return to the nebular spectrum : although the main problem has 

 been cleared up in the way described, it would still be an important step 

 to imitate the spectrum in the laboratory, not so much to confirm the 

 origin of the lines as to get direct information about the conditions under 

 which they may be excited. No success has yet been obtained in this 

 direction, but it is fairly clear how the attempt should be made. We 

 must have conditions capable of exciting the lines of doubly ionised oxygen 

 and attempt to work in a large volume at high rarefaction. 



A large volume and high rarefaction (rarity of collisions) is suggested 

 by the nebular conditions, and was plausibly held by Bowen to be an 

 essential. It must be allowed, however, that such experimental evidence 

 as we have at present on passage downwards from metastable states does 

 not definitely point in this direction. 



