SOUBCES OF 



result m any higher degree of incandescence of the luminous 

 crater. The condition of the carbon is already such that it is 

 at the point of volatilisation, so that increase of current cannot 

 raise it to a higher state of incandescence. But, the greater 

 the current, the higher the efficiency of the arc light from the 

 point of view of total light production, that is, increase of current 

 results in more than a proportionate increase in the area of 

 the incandescent crater, and consequently the amount of light 

 produced is greater. 



By far the simplest and most certain method of adjusting 

 the length of the arc and keeping it constant both in position 

 and condition is the use of a simple hand-feed arrangement. 

 As the result of experiments carried out in- collaboration with 

 Dr. T. A. B. Carver the writer adopted and has since used 

 constantly a simple hand-feed method of arc regulation, and 

 in spite of the improvements made in automatic lamps some 

 such device is to be preferred. 



Figs. 31 a and b illustrate the method of adjustment. 

 The carbons are held in V-clamps sliding upon two vertical 

 rods, and their relative positions can be varied by turn- 

 ing the right- and left-handed screw which is controlled by 

 a milled head. The position of the whole arc, that is of both 

 the carbons, can be varied without altering the length of arc 

 by turning the screw B, as this engages in the bracket E, 

 causing it to move up and down the two vertical rods C and D. 

 The lamp may be held by any convenient method, but is shown 

 mounted on a boss sliding on a tube. Its inclination can be 

 altered by loosening the screw F, which allows any required 

 inclination to be obtained. As the figures show, the lamp 

 can be carried either on a vertical or on a horizontal rod, so that 

 it is easily adaptable to any conditions under which it may be 

 required to work. With this arrangement it is quite simple 

 to feed the arc so that it maintains its truly central position 

 in relation to the microscope and constant distance between 

 the carbons themselves. 



Any other simple and exact method of feeding the carbons 

 may be adopted, and Fig. 32 shows another form of mechanism. 

 Here the two clamps holding the carbons are supported in a 

 centre casting, and are actuated by an ordinary rack-and- 

 pinion motion, this latter being controlled by the milled head B. 



