Polarization of scattered sunlight in deep water 429 



plane of incident light polarization. Hence the e vector cannot be recognized directly 

 by inspection without rotating the instrument. 



Another important property of the analyzer described lies in its high scnsiiiviiy. 

 Small percentages of polarization of the order of 5% may readily be delected. The 

 instrument has, therefore, proved useful in addition to its present application in 

 monitoring the illumination of experimental apparatus where polarization of the 

 light source or stray polarized light in the system had to be avoided (Wahrman 

 1950; 1954 b). 



It is of interest that, although interference patterns provide a standard means of 

 structural analysis in crystallography, their application in the present instance reverses 

 the known and unknown factors in the optical parameters. In the measurements 

 here described it is not the crystal plate whose characteristics are to be determined 

 but rather the polarization of the deep-sea illumination which traverses it. 



A further convenient property of the polarization analyzer used is the fact that the 

 interference figure is optically located at infinity. Consequently the instrument may 

 be fastened directly over the lens of a camera focussed at its principle focus and a 

 record obtained which indicates whether or not polarized light is present or not, even 

 in small amounts. Its type and plane, if it is linearly polarized, will likewise be recorded 

 with precision. 



A complication is introduced for deep-sea work, however, by the need for enclosing 

 the camera in a water-proof case whose window is almost certain to exhibit strain 

 birefringence when subjected to pressure. Such an effect was actually observed in 

 some preliminary photographs taken at about 30 m depth with a hand-held 

 camera.* 



In that instance the camera with the analyzer attached over the lens was protected in 

 a transparent plastic box. This was apparently birefringent when the pictures were 

 taken since the figure recorded showed elliptical polarization in the horizontal light 

 pattern. Such has never been observed visually under water in level lines of sight 

 (Waterman, 1954 a) nor was it ever found in deep water photographs where the 

 possibility of strain birefringence had been eliminated. 



To obviate this difficulty in the present instance the analyzer was mounted outside 

 the pressure case enclosing the camera. Both front and back of the optical sandwich 

 involved were thus exposed to equal pressures. No detectable effect of the hydrostatic 

 pressure itself (about 20 atmospheres at the deepest station) was ever observed. Of 

 course, any strain birefringence arising in the window through which the camera 

 looked out of its case would have no effect in this situation since it occurred between 



the analyzer and the lens. 



The camera employed was a 35 mm Robot with an f/2-8 Zeiss tessar lens. It Ns-as 

 arranged to be triggered by a solenoid controlled from the ship. After each exposure 

 the film was automatically advanced by the camera mechanism so that 15-20 pictures 

 could be taken on a single lowering of the apparatus. The Robot was mounted in a 

 Ewing-type deep-sea camera case sealed with an 0-ring. The lens was aligned with a 

 plane glass window outside of which the polarization analyzer was attached as already 

 described. A two-conductor insulated cable ran from the solenoid operating the 

 camera shutter to the deck laboratory o^ Atlantis where the exposures were controlled. 



