On this particular occasion the 9 meter reading was unusually near the bottom. 

 Had the bottom reading been taken at 8.5 meters, the presence of the colder 

 water m.lght have escaped notice. It is entirely possible that similar thin 

 ritrata of cold water escaped notice on other dates, when the bottom tempera- 

 ture was taken one meter above the bottom. But, obviously, little importance 

 can be attached to strata of such thickness, especially when the temperature 

 gradient is no greater than in the case cited. In the remaining fifteen 

 series the top and bottom temperatures were identical on four occasions, 

 and the maximum difference observed was 0.95" C. 



That the almost corrplete absence of stratification at Station l58 was 

 not a local peculiarity is shown by a summary of the temperature record at 

 Station 8F (North Passage) as given in Table 9- Of the 12 series taken in 

 1929, only two show a marked temperature gradient: those of June 17 and 

 June 2U. On June 17 the change from surface to bottom was gradual and there 

 was no thermocline as it is commonly defined, that is, a stratum in which the 

 change is at least one degree centigrade per meter. On June 2U there was a 

 thermocline between 8 and 10 meters, where there was a temperature difference 

 of 3.145°. It is probable that stratification was quite general at this time, 

 for on the following day a thermocline was found between 8 and 9 meters at 

 Station 68 (Niagara Reef), and between 7 and 8 meters at Stations 75 (West 

 Sister) and 131^ (Middle Sister). In the three cases just mentioned, the 

 gradient was less than 2' per meter. 



Another period of stratification occurred earlier in the season, as 

 indicated by a vertical series of readings taken at Station 60 (Gibraltar 

 Island) on May 30. On that date the surface temperature was 21.75* and the 

 bottom was 12.75', or a change of 9.0' in 7 meters. Itiere were two transi- 

 tion zones present, one in the upper 1.5 meters and the second in the stratum 

 between 3 and 5 meters, each underlain by a stratum in which the gradient was 

 less marked. Judging from tenperatures taken at other stations before May 30, 

 and from meteorological data for May 31 to June 3, when the next water temp- 

 eratures were taken, the entire period of thermocline formation and destruc- 

 tion lasted only 6 days. 



The temperature record at Station l58 for 1930 (Table 10) is almost as 

 free from evidence of stratification as the record of 1929. On only two 

 dates (May 6 and June 3) was there a temperature gradient great enough to be 

 termed a thermocline. On May 6 it was located in the stratum between 7 and 

 8 meters, and the gradient was 1.14* for that one meter stratum. On June 3 

 there was a thermocline between 5 and 6 meters, but the gradient was only 

 1.0*. On the remaining twelve dates there were only insignificant differences 

 between the surface and bottom. 



Likewise, the record at Station 8F for 1930 shows only two examples of 

 -:itratification (Table 11) . Qn May 8 there was a thermocline with a gradient 

 'of 1.5* between 8 and 9 meters. On June 25 the thermocline was located be- 

 tween 10 and 11 meters, and the gradient was again 1.5*. The first of these 

 two instances belongs to the same period of thermocline formation as the one 



U5 



