BULLETIN 55, UNITED STATES NATIONAL MUSEUM. 



The following table presents an abstract of the records of bottom 

 temperatures : 



Depths. 



Less than 500 fathoms . 



500 to 600 fathoms 



600 to 700 fathoms 



700 to 800 fathoms 



800 to 900 fathoms 



900 to 1,000 fathoms . . . 

 1,000 to 1,500 fathoms.. 

 1, 500 to 2,000 fathoms. - 

 2,000 to 3,000 fathoms.. 

 3,000 to 4,000 fathoms.. 

 4,000 to 5,000 fathoms. . 

 5,070 and 5,101 fathoms 



Number 

 of obser- 

 vations. 



High. 



1 

 1 

 3 



5 



7 



3 



42 



83 



«266 



&188 



3 



2 



40.5 



38.6 



41.1 



37 



38 



39 



36 



36.3 



35.6 



36 



Low. 



38.3 



36 



36.7 



36 



35 



35 



34.2 



34 



35.4 



35.9 



Average. 



43.7 



39.8 



39.4 



37.3 



38 



36.4 



35.46 



35.31 



35.17 



35.22 



35.50 



35.95 



nl6 records thrown out. 



b 10 records thrown out. 



The high temperature average, between 800 and 900 fathoms, is due to 

 the exceptional record of 41°.! at station 1225, in immediate proximity 

 to the volcanic island of Oshima or Vries Island, at the entrance 

 to the Gulf of Tokyo; also two records of 39° at stations 1569 and 

 1570, on the summit of a high peak or ridge about 450 miles to the 

 eastward of the island of Guam. The average of the other four 

 records is 36^.97. 



In the series of observations at depths between 1,000 and 1,500 

 fathoms there is record of 38° at station 1678, and 37°. 3 at the adjoin- 

 ing station 1677. These two stations are on one of the peaks of the 

 mountain range in midocean between Midway and Guam. There are 

 no other records of temperature above 37° at these depths. 



Only two stations between the 1,500 and 2,000 fathom line record 

 temperature above 37°, namely: Stations 1000, 39°. 3 and 1009, 37°. 3, 

 about 60 and 120 miles, respectively, to the northward of Guam. There 

 is probability of error in one or both of these observations. 



In making up the average of temperatures between 2,000 and 3,000 

 fathoms, 16 of the 266 observations have been omitted from the 

 calculations. In some of these cases "incorrect" is noted on the 

 original record; in others, the probability of instrumental or clerical 

 error is so much greater than the probability of existence of local 

 conditions capable of producing such deviations from the normal 

 range of temperature as to justify their exclusion. The omissions are 

 stations 131 (44°), 138 (51°.7), 140 (44°.8), 232 (39°.4), 243 (38°.2), 

 477 (38°.8), 479 (44°.5), 719 (67°.6), 722 (67°), 723 (67°), 962 (37°), 

 1508 (33°), 1511 (33°), 1512 (34°), 1513 (34°), 1514 (34°). The last five 

 of these rejected observations were taken b}^ a thermometer concern- 

 ing which it is noted: " Correction not known." This thermometer 



