26 METEOROLOGICAL RESULTS OF LAST CRUISE OF CARNEGIE 
5° to 15° south (38 days, November and January - April) 
show: latitudes 15° - 25° south, c 3=0°034C, $3=331°79; 
latitudes 5° - 15° south, c 3=0°047C; $3=32673. 
For equatorial and tropical regions, between lati- 
tudes +20°, the amplitude of the 8-hour temperature os- 
cillation, as computed from the Carnegie data, is about 
one-tenth the amplitude of the 12-hour oscillation. 
There appears to be little regularity in the differences 
of phase between the two temperature waves; the maxi- 
mum pressure varies from 3.8 hours earlier to 2.8 hours 
later than the maximum temperature. Between latitudes 
+20°, the time of maximum pressure averages 2.3 hours 
earlier, a result which compares favorably with the fig- 
ures of Bartels [29] for Potsdam, where the pressure 
maximum in summer averaged 2.0 hours earlier than 
the temperature maximum. 
Finally, brief mention should be made of the 6-hour 
variation in temperature. The only extensive work on 
this variation has been conducted by Pramanik [18], who 
concludes that the phase, $4, is fairly uniform with a 
slight seasonal variation from a mean of 204° in sum- 
mer to 250° in winter. The mean annual amplitude, c4, 
is a maximum (0°23) at latitudes +25° (approximately), 
decreasing to 0°10 near the equator and to 0°05 at lati- 
tudes +50°. He found the amplitude smaller in summer 
than in winter and smaller at coastal stations than in- 
land. 
The irregularity of the Carnegie Fourier quantities 
for this 6-hour term precludes any detailed comparison 
with Pramanik’s data. Figure 21, however, serves to 
indicate one conspicuous difference between the Carne- 
gie values for c4 and Pramanik’s values for several 
stations in corresponding latitudes also under summer 
conditions [18, p. 55], that is, the small amplitude of the 
term from the Carnegie observations when compared 
with the large amplitude determined for these coastal 
stations by Pramanik. 
CONCLUSION 
The Carnegie data show that both spatial and time 
variations in air temperatures over the ocean are usual- 
ly small. Owing to the slightness of such variations, 
however, it becomes increasingly evident that the accu- 
racy of air-temperature measurements on board ship 
must be increased before detailed studies of these vari- 
ations can be undertaken. Quite frequently the air tem- 
peratures recorded on the Carnegie show marked dis- 
crepancies, due, no doubt, to local heating and cooling of 
the vessel and thermometers; and, not many adequate 
temperature analyses could be made in view of the con- 
sequent inaccuracy of the data. It is to be hoped that 
greater attention will be given on future expeditions to 
the details of air-temperature measurement at sea, and 
that these Carnegie results may serve to stimulate in- 
vestigation and experimentation in this field. 
