on the filament rather than on the bulb wall, as with a normal 

 incandescent lamp. This results in very nearly 100 per cent 

 lumen maintenance and greatly increased lamp efficiency. 



Maintenance of the iodine cycle, W+2I = WI 3 , depends upon 

 the maintenance of a 3000°C filament temperature and a wall 

 temperature of at least 6 00°C. This high-temperature 

 requirement is a definite advantage in a bulb that must be 

 sealed in a heavy casing where there is no ventilation, since 

 it is very difficult to radiate the heat through the heavy 

 pressure casing as would be required with a normal incan- 

 descent lamp. This factor, coupled with the high luminous 

 efficiency of the bulb, made it a natural choice for this 

 application. 



The first item in the development of the pressure casing was 

 the window. As with nearly all "window" applications for 

 the bathyscaph, Plexiglas was selected as the best material. 

 While quartz, which is commonly used in pressure applica- 

 tions, might seem to be the logical choice, it has a tendency 

 to fail because of even very minor flaws or scratches. A 

 long-continuing program of tests on Plexiglas, 

 beginning with Prof. Auguste Piccard's tests on the Plexiglas 

 windows used in the bathyscaph, have shown that this mater- 

 ial can be used safely in all pressure applications provided 

 certain design criteria are followed. Plexiglas also has the 

 great advantage of equalizing its own stresses under pressure, 

 A scratch, which on the face of a quartz window would un- 

 doubtedly cause failure under pressure, would do nothing 

 more than slightly retard the transmission of light in Plex- 

 iglas. 



The first test casing was designed and constructed of steel. 

 In the design of this case, a fine mesh brass screen was 

 used between the lamp and the Plexiglas. Its purpose was 

 to carry the heat generated by the lamp to the walls of the 

 casing. This prevented the Plexiglas from flowing under 

 the high temperature imposed upon it by the lamp. The 

 casing worked well under full ocean pressure; however, 

 the fine mesh screen cut down the light transmission. It 

 was therefore believed that under conditions of long- duration 

 use, the heat would not be carried away sufficiently to pre- 

 vent the plastic from flowing. 



The second case was designed with a water barrier between 

 the lamp and the Plexiglas. This water barrier was main- 

 tained by a piece of optical glass (fig. 4). The "production" 

 model casing that was finally selected utilized the water 

 screen, but was constructed of aluminum instead of steel. 



