1162 
Journal of Agricultural Research 
Vol. XXX, No. 12 
recorded as partly cloudy, we have 158, 
155, and 145 days of the 214-dav 
period as days of sunshine in the three 
years. Expressed as percentages, these 
equal 73.8, 72.4, and 67.7 per cent of 
the total number of days in the period. 
It must be emphasized that in making 
this comparison an arbitrary valuation 
has been assigned to the sunshiny 
hours of partly cloudy days, and that 
the figures arrived at must not be 
regarded as exact mathematical state¬ 
ments of what occurred in the three 
seasons in question. It is clear from 
the data in Table VI, however, that 
for the period covered by this work 
the year 1920 had the smallest number 
of clear days and the largest number 
of partly cloudy or wholly cloudy 
days during the growing season, and 
month occurring on the 26th, 28th, and 
29th. While the only precipitation in 
October occurred on the 28th the period 
between the 1st and 16th was remark¬ 
able for the nightly dense fogs which 
usually persisted till midafternoon and 
broke away to recur shortly after night¬ 
fall. The temperatures during the 
nights and forenoons were low, and 
though there were short periods of sun¬ 
shine in the afternoon during wdiich a 
fairly high maximum was attained the 
ripening of the fruit was slowed down 
greatly and some varieties did not reach 
normal coloration. The harvest ex¬ 
tended from September 18 to October 
17, at least a week longer than in any 
other year. 
The persistence of warm weather 
through November delayed the ap- 
Fig. 6 .—Maximum, minimum, and mean temperatures at Vineland, N. J., in 1920, with 50-year 
average for comparison 
that consequently the sunshine re¬ 
ceived was minimum in amount as 
compared with the other years of the 
period. While the fruit of most of 
the varieties remained on the vines a 
week to two weeks after the end of 
September, the season being exception¬ 
ally late, this did not materially 
increase the amount of sunshine re¬ 
ceived for a reason stated in the next 
paragraph. 
The situation which actually existed 
during the period September 15 to Octo¬ 
ber 15 is not indicated by the monthly 
summary of weather data in the graph. 
The period September 13 to 23 was 
rainless, and ripening of the earlier vari¬ 
eties proceeded normally. A period of 
cloudy, sunless weather began on the 
24th and continued through the remain¬ 
der of the month, more than three- 
fourths of the precipitation for the 
pearance of killing frost until Novem¬ 
ber 12. There was an excess of rain¬ 
fall in both November and December 
with total absence of snow and the 
usual December temperatures. 
Climatic Record for 1921 
The outstanding features of the 
climatic record for 1921 are the almost 
total absence of typical New Jersey 
winter weather and the great deficiency 
of precipitation throughout the State 
(10.07 inches at Vineland) (figs. 8 and 
9). The mean temperatures at Vine- 
land for the four months January to 
April, inclusive, were 4.6°, 5.4°, 14.4°, 
and 8.6°, respectively, above the 50- 
year average, the minimum tempera¬ 
ture recorded in March being 1.7° 
above the normal mean temperature 
for the month. This highly abnormal 
