42 Papers from the Department of Marine Biology, 



CALIBRATION OF AN IMPROVED INDICATOR METHOD FOR USE ON 

 SHIPBOARD AND IN POORLY EQUIPPED LABORATORIES. 



After the excellent work of Sorensen and his associates on the 

 indicator method it might be supposed that the last word had been 

 spoken on this subject. But there are several disadvantages in 

 Sorensen's method. The solutions have to be exposed to the air during 

 the determinations and are placed in such a position that the operator's 

 breath might blow directly into them. (This has been obviated in 

 our method by sealing the tubes by fusing the glass, as done by Rown- 

 tree and his associates.) The salt-content of Sorensen's buffer mixtures 

 is much less than in sea-water, and a large salt correction had to be 

 applied, and no new indicators could be used until their salt error was 

 determined. This has been obviated by calibrating new buffer mix- 

 tures of the same salt-content as sea- water. The use of a colorimeter 

 that brings the centers of the tubes together in a sharp line makes it 

 possible to compare the sealed tubes from the side even more accurately 

 than open tubes viewed from above. The sealed tubes are gradually 

 changed by solution of even the most resistant glass, but this change 

 has been reduced by increasing the volume-surface ratio as well as 

 the buffer and NaCl concentration. The selection of the brilliant 

 sulfophthalein indicators, brought to our notice by the work of Lubs 

 and Clark, greatly facilitates the determinations. We found that the 

 salt error of all of these indicators so far calibrated is the same ; at least 

 no distinction could be made within the range that we had to use. 



Although we made the salt-content of the buffer mixtures the same 

 as sea-water, it is sometimes convenient to apply a slight salt correction 

 for the shght variation in the salinity of ocean-water rather than 

 carry a separate set of tubes for each salinity. But the magnitude of 

 this correction seldom exceeds one-sixth of the correction that had to be 

 applied to Sorensen's buffer mixtures. In our final set of tubes, a 

 slight compromise was made in order to satisfy all of the principles set 

 forth at once. Borax is the best buffer and is not very soluble; its 

 solubility is reduced by the presence of salt. Since it was desired to 

 have as high a concentration of borax as possible, the salinity of the 

 buffer mixtures was made to equal that of sea-water of about the 

 lowest salinity ordinarily met with, and the labels of the tubes were 

 changed so as to correct for salt error of sea-water of the mean salinity 

 that is to be encountered in an investigation, and directions were given 

 for correcting for variation in saUnity. Unless otherwise specified the 

 tubes are labeled for a salinity of 31.25 grams per kilogram and with 

 these tubes salinity variation within the ordinary range of the larger 

 oceans causes slight errors that may be allowed for. 



The buffer mixtures were made of different proportions of two stock 

 solutions. In making a set of tubes, the indicator was selected and 

 added to portions of the two stock solutions in two automatic burettes 



