of carbon per hour at 1,000 ft.-c,, the use of a 

 foot candle notation is not strictly correct, 

 since the emission characteristics of the lamps 

 are rather different from the relative spectral 

 sensitivity of the human eye. Furthermore, the 

 use of illumination units implies wrongly that 

 the plants themselves have a spectral response 

 comparable to the human eye. 



A more appropriate measure is irradiance 

 {i.e., ergs per second per square centimeter). 

 Illumination may be converted to irradiance 

 with the following conversion factor: 0,33 x 

 10"5 erg/sec./cm.2 jg approximately equiva- 

 lent to 1 ft.-c. This factor was obtained by 

 making a direct comparison of the foot-candle 

 output (with a WestonlUumination Meter) of the 

 incubator light source, in the range 30 to 90 

 ft.-c, and the energy as measured with a 

 vacuum thermopile and auxiliary equipment. 

 This value is almost identical with that 

 given by Strickland (1958). Most of the very 

 near infrared (700-800 m/i ) was filtered out by 

 placing a cell (3.2- cm. path) containing copper 

 sulfate solution (10 g. CuSo4«5H20 per liter in 

 0,5 percent H2SO4) between the thermopile and 

 source. Longer wavelengths are not emitted in 

 appreciable amounts from fluorescent lamps 

 and thus would not introduce a sizeable error 

 in the thermopile measurements (Withrow and 

 Withrow, 1956, figs, 3-15). This solution was 

 not used in the foot-candle measurement, since 

 the photronic cell of the Weston Illumination 

 Meter is not sensitive to wavelengths greater 

 than 750 m^ . 



The illumination level in the bath was 

 adjusted to 1,000 ft.-c, by varying the lamp 

 ballast voltage, and maintained at nearly this 

 value (+2 percent) during the incubation period. 

 As the fluorescent lamps aged with use, the 

 light output dropped. The output of aging 

 lamps could be raised to something in excess 

 of 1,000 ft.-c, by switching in a second ballast 

 to each lamp. The final adjustment to 1,000 

 ft,-c. was again made by adjusting a Variac 

 which controlled lamp ballast voltage. 



The equipment used during this investigation 

 for surface and deck incubation has undergone 

 several changes, and somewhat different 

 methods have been followed on different 

 cruises. 



The first method used was to trail surface 

 samples astern of the vessel immediately 

 below the sea surface. The samples were 

 placed in a brass frame holder designed 

 to shade the bottles as little as possible. 

 This technique was used on Expeditions 

 EASTROPIC and SCOPE. 



A more elaborate but similar piece of equip- 

 ment was used on Expedition SCOT. This trail- 

 ing incubator consisted of a plastic tube into 

 which the sample bottles were inserted. The 

 tube gave enough clearance to allow neutral 

 photographic filters to be placed around a 

 number of the bottles, which were separated 



from each other by opaque spacers. Samples 

 for incubation were collected at optical depths 

 corresponding to the transmittance of the 

 neutral filter series and covered with the 

 appropriate filter during the incubation period. 

 This unit was not entirely satisfactory because 

 condensation of water vapor in the tube during 

 towing damaged the neutral filters and changed 

 their transmission characteristics. 



Considerable care was used to prevent the 

 trailing bottle assembly fronn bouncing on 

 surface waves and in and out of the surface 

 film. This control was achieved on different 

 vessels by different handling methods which 

 consisted of varying the length of towing line, 

 or weighting the towing line with chains to 

 keep the samples submerged about 0.5 to 1 m,, 

 or both. 



A deck incubator replaced the trailing bottle 

 after the SCOT Expedition. This unit was 

 similar to that used by Berge' (1958). The 

 incubator was cooled with surface sea water, 

 and the compartments in it could be covered 

 with neutral photographic filters of various 

 densities. As previously, irradiance values 

 were determined with the irradiance meter 

 and samples for incubation collected at depths 

 corresponding to the transmittance of the 

 graded series of neutral photographic filters. 



In situ productivity measurements were 

 obtained whenever vessel time could be spared. 

 Collection and inoculation techniques were 

 identical with those described above. The 

 sample bottles were fastened on a weighted 

 cotton line at intervals corresponding to their 

 collecting depths and lowered into the sea. The 

 line was supported at the surface by plastic or 

 glass floats, which also supported a bamboo pole 

 bearing a radar reflector and an identification 

 flag. After launching the assembly, the vessel 

 left the immediate vicinity and proceeded with 

 other observations. At the end of the incuba- 

 tion period, either local apparent noon or sun- 

 set, depending upon whether the samples had 

 been launched within a half-hour of sunrise 

 or local apparent noon, the buoy was retrieved 

 and the samples were filtered. 



Although this incubation technique is sup- 

 posed to duplicate the natural system closely, 

 certain difficulties prevent attainment of nat- 

 ural in situ photosynthetic rates. Shock from 

 temperature, pressure, or light changes may 

 result from bringing deep samples to the 

 surface for inoculation with C^^ before the in 

 situ incubation can proceed. The effects of 

 these changes have not been evaluated in the 

 present study, nor do other investigators ap- 

 pear to have exannined this problen-i. 



The technique as described presents an 

 additional problem. The surface float system, 

 even when glass floats are used, is nearly 

 opaque; it is doubtful that estimates of 

 surface sample rates can be considered 

 reliable. 



23 



