4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 60 



As my time was too fully occupied in other work to permit of 

 my devoting - my personal attention to it, the greater part of the 

 actual determination of the densities was done by Mr. Leonard M. 

 Tongue, the clerk of the ship, under my supervision. Mr. Tongue's 

 familiarity with the instruments and his conscientious devotion to 

 the scientific work of the ship combined to make his calculations 

 reliable within a comparatively small limit of personal error. 



In addition to surface specimens, samples of water were, so far 

 as possible, taken at the same time from the circulating pump the 

 intake of which is about four feet below the surface. This water 

 was drawn from the tap in the laboratory, enough water being first 

 allowed to run off entirely to clear the pipe from the laboratory to 

 the sea. It was not believed that the modification of the water on 

 its passage through this pipe, which was constantly in use, could be 

 sufficient to be detected by the method used in the determination of 

 the specific gravities. 



Many of the salinity determinations taken at the surface during 

 this cruise are included, without correction or comment, in the dredg- 

 ing and hydrographic records of the cruise (Bureau of Fisheries 

 Document No. 621, 1907, pp. 1-50). It was necessary in distributing 

 our zoological material for study to provide each of our collaborators 

 at the same time with the data of each dredge haul. Immediately 

 upon our return to San Francisco, therefore, Mr. Tongue compiled 

 these records from our undigested notes taken at sea. 



THE LIMITS OF ERROR IN THE OBSERVATIONS 

 There were fifty observations taken on the water from the circu- 

 lating pump ; in three instances these readings were identical with 

 those taken from water dipped from the surface (Nos. 24, 64 and 

 73) ; in twenty-four instances they were greater, as follows: 



and in twenty-three instances less, as follows 



