32 METEOROLOGICAL RESULTS OF LAST CRUISE OF CARNEGIE 
Table 36. Frequency in days of unperiodic daily amplitude of sea-surface 
temperature, Carnegie, 1928-29, and on Gauss, 1901-03 

Vessel 

Carnegie 55 142 66 
Gauss 104 109 54 
Table 37. Diurnal amplitude of sea-surface 
temperature according to ranges in 
latitude, Carnegie, 1928-29 

Range in Diurnal Range in Diurnal 
latitude amplitude latitude amplitude 
° ° “Ye ° ° °6E 
>45 N 12 5N- 5S 1.0 
45 N-35 N 2.3 58-155 0.9 
35 N-25 N 1.2 15 §-25 S 0.7 
25 N-15 N 0.8 25 §-35 S 1.1 
15 N-5N 0.7 35 S-45 § 1.6 
INE orn oc 2's EE ict-fale ocrelote 1.15 

interesting to consider that the minima for sea-surface 
and air temperature occur at the same hour, but that 
there is a lag of two hours in the sea-surface maximum. 
Variation of the Diurnal Amplitude 
of Sea-Surface Temperature with Latitude 
Table 37 shows the variation in the daily range of 
sea-surface temperature with latitude. It appears that 
the amplitude generally increases with latitude, although 
it may be noted that the amplitude is somewhat greater 
at the equator than between latitudes +5° to +25°, and 
that the amplitude at mean latitude 40° north is unusual- 
ly large. The explanation for the maximum at the equa- 
tor appears to be that the amplitude within the range of 
latitude +5° is greatly influenced by the number of ob- 
servations made on board the Carnegie in the vicinity of 
the Galapagos Islands, where noncyclic changes in sea- 
surface temperatures were large. Similarly, the ampli- 
tude between latitudes 35° and 45° north appears to have 
been greatly affected by observations made within the 
California and Kuroshio currents, where, again, non- 
cyclic effects were important. The curve produced by 
the data in table 37 would probably be more regular if 
more data were used in computing the means. 
The elimination of all days with a range of sea-sur- 
face temperature greater than 3°, however, does not 
appreciably affect the results. 
Effect of the State of the Sea on the Diurnal Variation 
of Sea-Surface Temperature 
Since mixing of the surface sea waters should tend 
to reduce the daily amplitude of sea-surface tempera- 
Amplitude in °C 

26 25 16 
24 30 15 
330 
336 
tures, a comparison of the differences between the un- 
periodic amplitude for days with moderate-to-rough sea 
and for days with a smooth sea has been made for the 
purpose of determining to what extent this is true. 
Owing to the loss of the logbook of the Carnegie, it 
has been possible definitely to select days with 2 smooth 
or rough sea in only thirty-four cases. All these were 
days when the vessel was in tropical waters between 
latitudes +20°. The results give a value of 0°6 for the 
seventeen days with moderate-to-rough sea, and 1°4 for 
the seventeen days with smooth sea. Although it is real- 
ized that seventeen days of observation are not suffi- 
cient to determine this relationship satisfactorily, it is 
believed that the difference between the two sets of data 
(0°8) is sufficiently large to be conclusive. 
Van Riel [34] found that the change in the state of 
the sea from mean smooth to mean moderate reduced 
the daily range in temperature about 0°2. Information 
concerning the methods of determining the state of the 
sea for Van Riel’s data is not given, but it is assumed 
that these data were probably more accurate than the 
corresponding Carnegie data, and thus that his range is 
most nearly correct. 
Effect of Cloudiness on the Diurnal Variation 
of Sea-Surface Temperature 
It was considered especially interesting to observe 
the effect of cloudiness on the diurnal variation of sea- 
surface temperature, but, again, the loss of the Carne- 
gie logbook makes it difficult to separate the days into 
appropriate groups. Ten days have been chosen, how- 
ever, which were summarized in the log abstract as 
“clear days’’ (cloudiness less than 0.2) with wind force 
less than 4, Beaufort scale, and also ten cloudy days 
(cloudiness greater than 0.8) with wind force less than 
4. These twenty days were all in the tropical Pacific 
Ocean between latitudes +20°. 
The mean daily range of temperature for the ten 
cloudy days proves to be 0°66, and for the ten clear days, 
1°24. These values compare favorably with those de- 
termined by Schott [35] from observations on a sailing 
vessel and by Meinardus [4, p. 522] from the Gauss 
data (table 38). 
The mean 24-hour values, corrected for noncyclic 
change, were computed for each of the above ten-day 
groups and the results are shown in figure 25. The 
small periodic amplitude of 0°1 for the cloudy days is 
in decided contrast with the amplitude of 0°8 for the 
clear days. According to these curves, the minimum 
sea-surface temperature occurs at 07h on the clear 
days, and on the cloudy days at 02h. The maximum sea- 

