AS OBSEEVED IN TOEEICELLIAN AND OTHEE VACUA. 
139 
the tube becomes filled with a luminous discharge to within about one inch of the 
negative wire ; the stratifications appear gradually increasing in width, as the vacuum 
becomes more perfect ; and if care is taken to continue the pumping, so as to prevent 
air being introduced, the tube can be sealed without the stratifications showing the 
slightest appearance of any redness. Dr. Feankland subsequently suggested to me a 
mode of exhausting these tubes, which efiectually prevents a possibility of the intro- 
duction of any air. 
64. In a vacuum-tube exhausted by the air-pump alone I have not been able to 
observe any approach to the cloud-liko stratifications which are always to be obtained in 
good Ton-icellian vacua. That they evidently do not depend either on the mercury or 
the tube being free from dirt, but are dependent on the perfection of the vacuum 
and the total absence of all trace of moisture, is evident ; as in one instance the cloud- 
like stratifications were very distinct at the first discharge, although in the tube the 
mercury was very dirty, attaching itself in black patches on the sides ; the dark band 
was nearly 10 inches in length while the mercury covered the negative wire. In another 
instance, during the process of the withdrawal of the mercury from a long tube, the 
discharges from the induction coil were continued, the "wire in the upper portion being 
positive and the mercury negative ; as the mercury descended the stratifications assumed 
the cloud-like appearance, each nearly one inch in length ; on examining the tube traces 
of moisture were observable adhering to the sides near the negative wire ; as soon as 
the mercury descended sufficiently down the tube, so as to expose the moisture to the 
vacuum, the cloud-like stratifications instantly changed to narrow bands. This tube was 
38 inches in length, and the wires 32 inches apart, the internal diameter being about 
one inch; the proportion of moisture to the total capacity of the tube must conse- 
quently have been very minute. 
65. Whenever a perceptible although minute residue of the gas or air remains in 
a Torricellian vacuum-tube, the stratifications are narrow and close ; and if further 
portions of air or of gas are subsequently introduced, the stratifications gradually 
close until the discharge assumes the usual appearance of the aurora, as when taken 
in an air-pump vacuum from the electric machine. In the best Torricellian vacua 
which I have been enabled to obtain, the stratifications invariably assume the character 
of a distinct and clearly defined cloud-like discharge ; these stratifications are generally 
more clearly defined towards the negative terminal, close to the dark band, than nearer 
to the positive. 
66. If care be taken in the preparation of Torricellian vacuum-tubes by Mr. Welsh’s 
process*, it is immaterial whether the original air-pump vacuum be from air or from 
either of the gases I have enumerated (62) ; the last trace of air or gas being by this 
process sealed off in the bulb, the discharge in the mercurial vacua will invariably pre- 
sent the cloud-like stratifications. If any appreciable trace of air or gas was permitted 
to remain, I always found, if air, the stratifications had a red tinge, which by degrees 
* Philosophical Transactions, 1856, p. 507. 
u2 
