PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. VI. 709 



The TROWBRIDGE explanation of the direct axial rotation may perhaps be equally applicable to our 

 tiory. He shows that if the ring be nearly of the same density throughout, the resulting planet must 

 1 ve a retrograde rotation like Uranus and Neptune. But if the particles are more closely packed near 

 . inner edge of the ring, so that the resulting planet would be formed much within the middle of its 

 dth, its axial rotation must be direct. 



Our results summarised above seem both simple and well fitted to aid in constructing a new and 

 s .isfactory cosmogonic hypothesis, based on experimental analogies (see experiments represented in fig. 255). 



137. We will here take the opportunity of mentioning some more recent experiments that have 

 ten made with the largest vacuum-box with a capacity of 1000 litres. We have already referred to 

 t:ir commencement. 



Our experiments, as might be expected, prove to be more and more interesting as we increase 

 t : scale on which they are performed. 



Fig. 263 a gives a good idea of the dimensions of the vacuum-box, and the various arrangements 

 f- the experiments. 



The glass walls of the box, each of which supports a pressure of about 7000 kg., are 46 mm. in 

 t rkness. No firm of makers would supply any thicker, but it was calculated that they should have 

 I en 50 mm. in order to be safe. The floor and roof of the box are constructed of brass. 



The largest cathode-globe employed is 36 cm. in diameter, and the maximal discharge-current has 

 hen about 400 milliamperes. Fig. 263 b shows how the rays in the magnetic equatorial plane may be 

 vry pronounced when the magnetism is weak and the discharge-current comparatively strong (150 milli- 

 a pcres). With a stronger discharge-current, a peculiar electric corona frequently occurs round the 

 c hode, sometimes with rays out from the polar regions, the whole thing having a striking resemblance 

 t> photographs of the sun's corona during an eclipse. 



If we desire to produce the phenomenon which we think may be regarded as analogous to Saturn's 

 r g, only i or 2 milliamperes is required, and the magnetisation of the cathode-globe must be some- 

 v at stronger than in the former experiment. 



Fig. 263 c shows powerful and characteristic spot-discharges from the magnetic cathode-globe. 



It will be necessary to give some information as to the way in which these disruptive discharges 

 ny best be brought about. 



With a polished metal globe like the cathode, disruptive discharges will not easily be formed. An 

 a lost continuous discharge with electric corona is then obtained, even if, as previously mentioned, the 

 s am of vaseline-oil be introduced into the box. 



If, on the contrary, the globe is cast and not polished, such disruptive discharges will nearly al- 

 wys occur; but the difficulty here is that the casting of so large a globe as the one here employed 

 n.-er yields a homogeneous result, so that the patches keep to certain parts of the globe, even when 

 tl: latter is not magnetised. 



The best way in which to obtain with certainty a continuous discharge, interrupted at definite inter- 

 s by powerful disruptive discharges, seems to be the following: 



The surface of the globe, after polishing, should be sand-blown and then painted over with a thin 

 ing of vaseline-oil, which is afterwards wiped off again. This painting over, which seems to be 

 antageous to the phenomena, is not necessary if the farthest corners of the vacuum-box are greased 

 h a little vaseline-oil before the box is exhausted. 



When a suitable vacuum has been obtained, a short discharge of about 10 minutes with a current- 

 s'sngth of 200 milliamperes will completely dry up the oil on the cathode. Without discharges the 

 gbe will remain oily for many days, even in a high vacuum produced by a GAEDE'S molecular pump. 



Birkeland. The Norwegian Aurora Polaris Expedition, 19021903. 90 



