AND STATICS UNDER THE INFLUENCE OF LIGHT. 
24 hours’ use—often even after much less—the candle power of the burner, either 
with the slits or pinholes, was no longer the same, and very often the slit or 
pinhole was already so much carbonised that the flame began to smoke ; on the 
contrary, pinholes with a larger diameter, giving 40-50 candle power the pair, 
require a very much longer time before they become carbonised and begin to smoke, 
though even their candle power also diminishes with time. Since the form and even 
the thickness of the flat flame changes continuously, we can only get a constant 
light by cutting out a piece from the middle of the flame for a certain time, screening 
all the rest of it. Since we cannot get one flat white flame, by means of several gas 
streams directed to the same point, of more than 60 or 70 candle power, the increase 
of thickness or of the size of the flame beyond the 70 candle power being always 
accompanied by the formation of smoke, we could thus, in the best circumstances, 
not get a constant flame of more than 20 candle power, a candle power not very 
different from standards already existing. On the other hand, it was impossible with 
several flat flames to get one light of great intensity in a small space, since the flames 
cannot be placed very near to one another (owing to the form of the burner), and a 
flame of 200 or 250 candle power (12-15 nipples) necessarily occupies a very large 
area. If we further consider that each of the flames is different from the others in 
size and form, and that the flame of any one burner soon changes in form and size, 
and that only a small part of each could be cut off securely by a screen so as to give 
a constant light for some time, it is evident that the number of lights or nipples 
which would be required for a 200-250 candle power light to remain constant would 
be about 40 or 50. Assuming that even the greatest care be taken in placing the 
burners and screens on as small an area as possible, still the burner would occupy too 
much space, and no point or line could be calculated from the different lights which 
could theoretically be assumed to be the point or line from which the total light 
was coming. 
The burner which was ultimately constructed free from these difficulties is that 
shown in the drawing (fig. 5). 
The wide tube A of the burner is divided into four narrower tubes as shown, 
leading at intervals of 90° into the channel of a hollow ring cut in the brass ring. 
In the hollow brass ring (AB) 12 small brass pieces containing capillary tubes were 
fixed, and in the end of the brass pieces nipples were fixed. Each nipple has one 
round pinhole of about 15 or 20 candle power; the holes are parallel to each other, 
and each gives a flame in the form of a straight thick line of a few millimetres 
diameter and of a few centimetres in height. Bound the brass ring a brass jacket 
is fixed, forming a hollow ring connected with the tubes r and v'. Water runs 
continuously through the hollow ring of the burner, which thus remains cool, in 
spite of the fact that flames of about 200-250 candle power are concentrated in a 
very small area. Using this arrangement of the burner, we find that the acetylene 
flame becomes perfectly pure and clear, and remains so for any length of time, even 
