326 Intelligence and Miscellaneous Articles. 



and nearer to the negative one. The latter begins to redden ; 

 at first the blue light continually grows fainter, then becomes in- 

 visible ; at least we cease to distinguish the shell that it forms 

 around the wire ; and if any trace of it remains, it is only a bluish tint 

 in the light due to the incandescence. When the wires are almost in 

 contact, especially if the finger is presssd lightly on the hammer of 

 the contact-breaker, the incandescence of the negative wire becomes 

 dazzling, and then there is no more appearance of the blue light. 



I was curious to know if it had really disappeared, or if it was 

 only concealed by the brilliancy of the wire when white-hot ; and I 

 thought that the now celebrated method by which we discover the 

 trace of the solar protuberances amongst the intenser rays of his 

 disk might be applied here. 



I made use of a vertical spectroscope by Duboscq. The slit 

 is vertical, and can be moved from one wire to the other along the 

 spark. The characteristics of the spectrum change according as we 

 view the brilliant point where the spark is detached from the posi- 

 tive wire, or the blue shell which envelopes the extremity of the 

 negative wire, or, finally, if that is incandescent, the red parts which 

 lie beyond the blue shell. 



We keep the slit upon the blue shell while the spark is too long 

 to admit of the wire becoming red. The spectrum is characterized 

 chiefly by a group of four green rays, a group of two rays placed be- 

 tween the green and the blue, a group of three violet rays, beyond 

 which we can see others of less brilliancy. 



As before, we gradually bring the positive wire nearer to the ne- 

 gative wire, which latter begins to redden. One would expect to see 

 a continuous spectrum ; and this in fact is what actually happens, if 

 we direct the slit towards the parts of the red wire which are 

 beyond the blue electric glow. We have then a continuous spectrum 

 which is worth noting, because we thus learn, without requiring 

 to light up the micrometric scale, that the violet rays given by the 

 blue light correspond nearly to the most refrangible extremity of this 

 continuous spectrum. Bringing back the slit to the extreme end 

 of the negative thread, we find again the streaked spectrum of 

 the blue light. The red in it becomes more brilliant in propor- 

 tion as the thread becomes more incandescent ; but the green, blue, 

 and violet rays still continue. But when the incandescence is very 

 intense, the green rays disappear, then the blue, and the spectrum is 

 continuous into the violet, but at the extremity of the violet we still 

 perceive the group of three violet rays, which become less distinct, 

 but mark their position until the thread begins to melt. The 

 ultra-violet rays have ceased to be visible. Thus the spectroscope 

 permits us in this case, as well as in the observation of the solar pro- 

 tuberances, to ascertain the presence of a feeble glow in the midst of 

 a light which to the direct vision is dazzling. — Comptes Rendus, 

 June 7, 1869. 



ON THE MEAN VELOCITY OF THE MOTION OF TRANSLATION OF 

 THE MOLECULES IN IMPERFECT GASES. BY M. P. BLASERNA. 



We are often led to inquire whence arise the deviations from 



