THE UNIVERSITY

OF ILLINOIS

LIBRARY

.2.

Y fl

N077CE: Return or renew all Library Materials! The Minimum Fee for
each Lost Book is $50.00.

The person charging this material is responsible for
its return to the library from which it was withdrawn
on or before the Latest Date stamped below.

Theft, mutilation, and underlining of books are reasons for discipli-
nary action and may result in dismissal from the University.
To renew call Telephone Center, 333-8400

UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN

L161— 0-1096

UNIVERSITY OF ILLINOIS

Agricultural Experiment Station

BULLETIN NO. 251

BLOOMING PERIODS OF APPLES

BY CHARLES S. CBANDALL

URBANA, ILLINOIS, MAY, 1924

CONTENTS OF BULLETIN No. 251

INTRODUCTION 11S

CHARACTER OF RECORDS 114

FLOWERING PERIODS VARIABLE H5

BLOOM RECORDS OF OTHER LOCALITIES H6

BLOOM RECORDS AT THE ILLINOIS STATION 120

Full Flowering Periods for All Varieties 120

Early Blooming Varieties 122

Late Blooming Varieties 123

Blooming Period of 1910 124

Comparison of the Periods for the Years 1904 and 1910 126

Varietal Flowering Periods 130

Amount of Bloom 131

Amount of Bloom and Distribution of Varieties into Time-Groups 139

Temperature and Distribution of Bloom 140

Inconstancy of Relative Lengths of Varietal Blooming Periods 141

SUMMARY . lii

BLOOMING PERIODS OF APPLES

BY CHARLES S. CEANDALL, CHIEF IN PLANT BREEDING IN HORTICULTURE

INTRODUCTION

The recording of periodical plant phenomena has received more at-
tention in recent years than formerly, and there is now abundant evi-
dence of general belief in the scientific and economic importance of
data of this character.

Flowering of fruit plants in spring marks, perhaps, the most
critical event in the production of a crop of fruit, and careful study
of circumstances occurring at that time, together with an effort to de-
termine and understand the influences that make for good or bad re-
sults, is deserving of attention, especially by those who make fruit
growing a business. Owners or managers of orchards would derive
satisfaction from annual records of flowering periods of the varieties
they grow, and compilation of numerous such records would aid
greatly in understanding the influences that cause wide variations
from year to year, but the Experiment Stations may rightly be ex-
pected to maintain more complete and continuous records of bloom-
ing phenomena than is usually possible for individuals because, once
started, the making of such records is regarded as a regular duty with
its proper place in the schedule of operations, and is maintained re-
gardless of changes in personnel. Many stations having fruit planta-
tions now maintain such records and each succeeding year adds to
their value. These records have been discussed in several valuable
bulletins, and short accounts of work done or in progress have ap-
peared in various station reports.

Blooming dates and periods are profoundly influenced by, in fact
appear to be in large measure dependent upon, meteorological condi-
tions, and, as these conditions vary extremely from year to year, long
continued records are necessary in order to supply a sound basis from
which to judge probable results under varying combinations of im-
portant influences. Any series of observations on plant events is of
most value for the locality where made, but in many features is of
value over wide districts, which possess somewhat similar soil and
climatic features. Surface features of the country, as a whole, are so
diverse, and the range of variation in all those elements that affect
growth, flowering, and fruiting of plants is so wide that records should
be kept at many different points by many observers. Out of the ag-
gregate of such records valuable light on the causes of many phenom-
ena now imperfectly understood, should eventually come.

113

114 BULLETIN No. 251 [May,

The purpose of this publication is to give a brief account of flower-
ing records for apples at the Illinois Station. The data to be con-
sidered include flowering dates and periods for 106 varieties having
records ranging from ten to sixteen years: the list of varieties in-
cludes most of those commonly grown in the state, and also a con-
siderable number of Russian varieties. The trees upon which obser-
vations were made were of nearly the same age, and were grown under
identical conditions. Many varieties were represented by more than
one tree, some by several trees, but the records to be mentioned are for
single individuals of each variety.

CHARACTER OF RECORDS

Records have been taken on forms prepared in advance and bound
in book form, one book for each year. Each page lists the trees of one
row, numbered consecutively according to position, and is ruled in five
columns with space allowed below for any remarks that seem desirable.
The numbers occupy the left-hand column, followed by the three date
columns: " opening of first flower," "full bloom," and "petals
falling." In the fifth column is recorded a general estimate of the
amount of bloom.

The record in the second column, except in rare instances, is a
simple statement of fact ; either a flower was open or was not open on
a given date, but cases of receptive stigmas protruding before petals
had expanded were sometimes encountered, and then the exact day of
first open flower was a matter of question. This condition was so rare,
however, with the varieties under consideration as to be negligible. In
some of the crab-like species of the genus, notably in forms of Mains
floribunda, Mains toringo, and Mains microcarpa, the maturity of
stigmas was so commonly in advance of expansion of petals that the
date of first open flower was often uncertain. Records of these crab-
like species, however, are all below the ten-year limit and are not to be
considered at this time.

Date of "full bloom" was often a matter of judgment and, hence
variable, especially in those seasons when flowers opened slowly be-
cause of storms and low temperatures. The third date, that of "petals
falling," not infrequently was a matter of judgment and subject to
fluctuation and error. Suppose that, following the opening of a few
flowers, a change in weather suspended plant activities for three, four,
or more days, this being followed by high temperatures and rapid
opening of flowers. Under such circumstances the petals of the first
flowers open would fall before the date of full bloom was reached. It
was assumed that the date to be recorded should apply to the mass of
flowers rather than to the few precocious individuals, but cases have
occurred in which fixing a satisfactory date was difficult.

BLOOMING PERIODS OF APPLES 115

The rule, sometimes used, of placing the date for " petals falling"
two days after full bloom may operate fairly well in some seasons, but
in other seasons cannot be used; thus in 1907, and again in 1908,
there were many instances in which the falling of petals was delayed
four, five, or more days following full bloom.

In the fifth column is placed an estimate of the amount of bloom.
Various systems of recording have been tried, including a decimal sys-
tem with figures ranging from 0 to 1, but whatever is recorded is an
estimate and fine distinctions are of no advantage. The plan now in
use recognizes five grades ; namely, no bloom, scant bloom, moderate
bloom, full bloom, and very full bloom, or as being more quickly re-
corded 0, 1, 2, 3, and 4. These distinctions serve every purpose. Un-
usual features exhibited by individuals have been separately recorded.

The ideal plan for recording flowering or any other phenomena of
plant life is for one individual to do the work in each and all years of
the series, but this is rarely possible. Seven men share the responsi-
bility of the nineteen years of phonological records at this station ; for
the sixteen years to be considered the field records were made by six
men: one made the record for five consecutive years, one for four
years, one for three years, one for two years, and two made each the
record of one year. An assistant has done the work for the last five
years, another assistant was in charge for the first two years ; for the
intervening years the work was done by senior or graduate students
under direction. Every effort was made to secure as accurate a record
as possible. Beginning before any flowers were open, each tree was
visited daily thruout the period of bloom, and during times of high
temperatures two and in some cases three visits were made each day.

FLOWERING PERIOD VARIABLE

For central Illinois the flowering period for apples is short ; in
some seasons very short. In some years heat waves occur that bring
trees into full bloom with astonishing rapidity. Flowers thus sud-
denly opened retain the ability to function for a very short time and
the whole flowering process is over almost before it is realized that
blooming time has come. On the other hand the spring season, in cer-
tain years, has been characterized by abnormally low temperatures, by
excessive rainfall, or by rapidly alternating periods of heat and cold,
so that flowering of trees has been very irregular and the season much
prolonged.

Presumably, the heat waves occurring in this region are more in-
tense, and hence more effective in pushing buds than in regions farther
north, and are less intense and less effective than in regions farther
south. However, comparisons of the length of flowering periods for
different regions cannot be safely based upon latitude alone, because
such local features as altitude, slope, soil, exposure to air currents, and

116 BULLETIN No. 251 [May,

proximity to bodies of water or belts of timber may be responsible
for wide variations.

BLOOM RECORDS OF OTHER LOCALITIES

Before entering upon detailed consideration of the bloom records
of this station, and for convenience of comparisons between dates and
periods here and elsewhere, brief mention may be made of a few of the
more important records published.

Virginia. — In 1905 Price1 of the Virginia Experiment Station, gave
blooming periods for 137 varieties of apples, including thirteen crab
varieties. The observations cover periods ranging from two to thirteen
years, but for about 64 percent of the varieties the observations extend
over nine or more years.

For each variety, the blooming period begins with the opening of
the first flower and ends two days after full bloom. The earliest com-
mencing date of bloom is April 21, the latest commencing date, May 5.
The earliest ending date is April 28, the latest, May 12. Aside from
the crabs, which, with the exception of Soulard, are among the earliest
to bloom, there are but eight of the late-blooming varieties the periods
of which do not, in some degree, overlap the periods of varieties bloom-
ing earlier. Considering all varieties, the flowering period — from the
opening of the first flower on the earliest blooming variety, to the end
of the period for the latest blooming variety — extends from April 21
to May 12, or twenty-two days. Individual periods range from five
days, as the average of nine years for Red June, to eleven days as the
average of three years for Seedless ; 3 varieties have six-day periods,
38 varieties have nine-day periods, and 5 varieties have ten-day
periods. The average period for all varieties is 8.14 days.

The latitude of the Virginia Station is 37° 15' ; the elevation
2,170 feet above sea level. A summary of meteorological records for
the twelve-year period 1893-1904 gives the mean annual temperature
as 51.5° ; mean maximum, 63.22° ; mean minimum, 40.04° ; the abso-
lute maximum, 96° ; and the absolute minimum, -13°. Mean annual
precipitation, 39.41 inches.

New York. — Flowering periods for a more extended list of varie-
ties, 278 in number, were given in March, 1908, by Hedrick2 of the
Geneva, New York, Station. Here the years of observation ranged
from three to six, with a little more than one-third of the varieties
included under the longer period.

The full flowering period extends from May 10 to May 29, or
twenty days; two days less than the corresponding period in Vir-
ginia, but the earliest flowering date in New York precedes the ending-

1 H. L. Price, Va. Agr. Exp. Sta. Bui. 155.

" U. P. Hedrick, N. Y. Agr. Exp. Sta. Bui. 299

1924} BLOOMING PERIODS OF APPLES 117

date of the period for the latest flowering variety in Virginia by only
two days. Varietal periods range from six days for each of 2 varie-
ties (Avery and Winter Citron), and seven days for 11 varieties to
twelve days for each of 5 varieties. Those varieties having average
blooming periods of ten days or more number 121 (or 43.52 percent
of the whole number). The average period for all varieties is 9.35
days.

In a later publication, Bulletin 407, issued in May, 1915, Hedrick
gives blooming records for 348 varieties for the five-year period 1910-
1914. About one-half of the varieties in this list appeared also in
the earlier list. In this bulletin, blooming periods for individual
varieties are not given, but the varieties are classified to show their
relation to each other as to earliness or lateness; they summarize as
follows :

Number Percent

Very early 10 2.87

Early 55 15.81

Midseason 239 68.68

Late 34 9.77

Very late 10 2.87

Of the blooming season Hedrick says, ' ' The blooming season of apples
at this Station averages twelve days. The shortest season of bloom
was seven days in 1913 and the longest season, eighteen days in 1910.
The first date of bloom for apples in 1912, an average year, was May
seventh. ' '

The latitude of the New York Station, as given in Bulletin 299,
is 42° 53' ; and the altitude, 525 feet above sea level. The Station is,
therefore, 5° 38' farther north, and has an altitude less by 1,645 feet
than that of the Virginia Station. From these differences in latitude
and altitude considerable differences in flowering time of apples should
be expected. But often there are local conditions of soil or climate
that have modifying influences as important as those of latitude and
altitude, and, from available data it does not seem possible to specify
the factors accountable for an average full flowering period two days
longer and an average individual variety period 1.21 days shorter
in Virginia than in New York.

The mean annual temperature at the New York Station, as given
in the report for 1917, ranged from 43.3°, in 1885, to 49.7° in 3913,
The average for the thirty-five year period is 47.4°. For the month
of May the mean for the thirty-five years is 56.5°, with the highest
mean, 64.09°, occurring in 1911, and the lowest mean, 51.3°, in 1907.
Annual rainfall has varied in the thirty-five years between 19.35
inches in 1899 and 38.69 inches in 1903 ; the average for the period
is 27.49 inches.

118 BULLETIN No. 251 [May,

Oregon. — Lewis and Vincent1 of the Oregon Station gave bloom-
ing records for ninety-five varieties of apples for the two years 1907
and 1908, with diagrams showing length of time in bloom for each
variety. The earliest commencing date is April 5 for Transcendent
Crab, the latest commencing date April 23 for Baldwin ; the earliest
date for petals falling is April 15 for Transcendent Crab ; the latest
for petals falling is May 2, and this date applies to eight varieties;
namely, Holland Peppin, Jonathan, Missouri Pippin, Yellow New-
town, Rome, Walbridge, Winesap, and York Imperial.

The days intervening between the earliest and latest beginning
dates number nineteen, and between the earliest and latest ending
dates, eighteen. The full flowering period April 5 to May 2 is twenty-
eight days. For individual varieties, the periods range from seven
days each for Martha Crab and Melon to fifteen days each for Tetofsky
and Whitney. Varieties having blooming periods ten or more days
in length number 78, or 82.1 percent of the total liumber ; the mean
length of period for all varieties is 11.31 days. Comparing this mean
of 11.31 days for all varieties for all years with the mean of 7.17 days
as found at the Illinois Station, it appears that the average flowering
period in Oregon is longer by more than 50 percent than in Illinois.
Varieties in the lists of the two stations are for the most part different,
there being only fourteen that appear in both lists, but, whether com-
parison is made between these fourteen varieties at the two stations
or between the entire lists, the difference in length of period remains
the same. As the Oregon record here considered covers but two years,
range of variation of periods in different years does not appear, but
in Bulletin 82 of November, 1904, Lake refers to variation in bloom-
ing as follows: "A very variable relation exists from year to year
not only between different varieties, but between the various phases
of development in the same variety, for example; in 1896 Delaware
Red and Domine were in full bloom seventeen days apart, but in 1897
they were in full blossom only two days apart." Further regarding
Fameuse he says, "The period from the opening of first blossoms to
the time when the tree is in full blossoms may vary from year to year
from three to fourteen days."

England. — Going outside our own country, there are available
certain English records of apple blooming periods that may be men-
tioned briefly. Blooming periods for apples in England, in general,
are longer than in this country and it does not appear that they are
subject to those extreme fluctuations in duration of periods in different
years that characterize records at many points in this country. Mean
temperatures in England are lower than at the Illinois Station; the

. I. Lewis and C. C. Vincent, Ore. Agr. Exp. Sta. Bui. 104.

1924} BLOOMING PERIODS OF APPLES 119

daily range is less, and the hot waves that are accountable for the
extremely short blooming periods often experienced here appear to
be wanting.

Of English blooming periods Director F. J. Chittenden1 of the
Horticultural Research Laboratory at Wisley in Surry says, "The
time of flowering of apples varies between rather wide limits, both
as to the commencement and as to the period over which the flowering
continues. It commences in southeast England about the third week
in April in early years and continues into the second week in June
in late years. During the four years 1908-1911, in which records
have been kept at Wisley, our earliest apple to open, Red Astrachan,
was in full bloom in 1910 on April 21 ; in 1911, May 2 ; in 1909,
May 3 ; and in 1908, May 7 ; while the most consistently late-flowering
variety, 'Royal Jubilee,' was in full flower in 1908 on May 23; in
1909, May 17; in 1910, May 20, and in 1911, May 19." "The period
during which one variety or other of apple was in full flower ( ignoring
the few days, about seven or eight, before, and the few, about eight
or nine, after full flowering time, during which some flowers were
open) was, in 1908, eighteen days (May 6 to May 23) ; in 1909, twenty
days (May 3 to May 22) ; in 1910, thirty-five days (April 21 to
May 25) ; and in 1911, eighteen days (May 2 to May 19)."

The same irregularities that characterize blooming records in
Illinois are found, also, in the English records ; that is to say, blooming
times and periods of varieties are not constant, they do not maintain
a definite order of blooming, nor can they be depended upon to flower
at a time having a definite relation to the flowering of other varieties.
The author quoted above arranges the blooming record of 168 varieties
in order from earliest to latest, and says of this arrangement: "It
is to be clearly understood that this 'average order of flowering'
represents only approximately what is likely to happen in any par-
ticular year, but it is believed that it is sufficiently accurate to serve
as a guide as to what varieties are most likely to be in flower at the
same time. One might choose any of a dozen above or below any
particular variety for planting with it to furnish pollen with the
certainty that any one of those varieties would sufficiently approximate
its full-flowering period to the one chosen. ' '

In a paper read before the Royal Horticultural Society2 entitled
"Observations on the Blossoming of Our Hardy Cultivated Fruits,"
Mr. Cecil H. Hooper gives a table showing average blooming periods
for the years 1908, 1909, and 1910 of 33 varieties of apples at Wye,
Kent, England. These average periods range from 12.66 days to 22
days; the average number of days in flower is 17.66. Each period

'Jour. Roy. Hort. Soc. 37 (1911-12) pp. 352-355.
a Jour. Roy. Hort. Soc. 36 (1910-1911) pp. 548-564.

120 BULLETIN No. 251 [May,

begins with the opening of the first flowers and ends when nearly all
petals have fallen. The full period for all varieties extends from
May 1 to June 9, or forty days. The number of days between open-
ing of the first flower and full bloom ranges for the varieties listed,
from four to ten, with an average of 7.42 days, and between full bloom
and completion of bloom from six to sixteen, with an average of 10.33
days. "In individual flowers the length of time from opening to fall
of petals is about seven days. ' '

Detailed consideration of blooming periods at the Illinois Sta-
tion may now be given. These records cover the years 1901 to 1916
inclusive, and include 106 varieties. Of these varieties 38 are recorded
as having bloomed in each of the sixteen years ; 24 varieties bloomed
in each of fifteen years ; 23 in each of fourteen years ; 11 in each of
thirteen years ; 5 in each of twelve years ; 3 in each of eleven years ;
and 2 in each of ten years.

Six other varieties of the older plantation have records of bloom
for numbers of years less than ten, but these are omitted, as are also
the records for trees planted in 1907 and 1908 and which now have
records extending from one to five years.

For the sixteen years, 1901 to 1916, the flowering periods ranged
from ten days to twenty-two days; the average was approximately
sixteen days. Eecords of two years, 1906 and 1911, approximate the
average; seven years had periods one to six days longer than the
average, and seven years had periods from one to six days shorter
than the average.

FULL FLOWERING PERIODS FOR ALL VARIETIES

By the term "full flowering period" is meant the period of time,
in days, between the first opening of a flower on the earliest blooming
variety and the falling of petals from flowers of the latest blooming
variety.

The earliest date recorded for commencement of bloom was April 2
in 1910, the latest date in the column "petals falling" was May 21,
1904. The extremes, then, are fifty days apart and the median date
would fall between April 26 and April 27. But there is here included
one year so abnormal as to stand entirely apart from all other years.
Reference is made to the year 1910, in which the first flowers opened
April 2 and petals of the latest flowering variety began falling April
19, one day later than the earliest opening of flowers in any other
year. If 1910 were omitted the full flowering period for all varieties

1924}

121

-0

os 1-1

• i—i

O

(M I-H fO

1 i-H

•^ »o

o

CM

r:

CO
(M C^IO

os

rH

t

os

Tj< U5

t^—i

(N-H 10

i— 1

l>

rH

CO

0

OS
r-<

co
<N r- 10

Tfl IO

i— 1
"3

rH

Tf<
rH

*>fO

1

H

~l
99

O
(N O5 TjH
1 1 1-H

e

*

3 2

5 CM

<o

O5

rH

rH
O

~

00 CO

CM — i CO

TP IO

rH

o -

rH

C5

00 •

r-(
03

O

os

C5
IM^HQO

•* -*1"1

00

l>

CO G

!

co

tX

tf

o>

OS

O5 CD
(M rH 00

Tfl IO

•O ""

•*

co

s

)

«

>

1C

w

0

5

CT>

T— 1 OS 1— 1

1 (M

Tj"O

02

§

, o
co

-

•*

H
r>

~

O

^H

O.-H

(N rH (N

1 1 (M

•*IC

Pn
o

O3

c-i

00
CM

0

e

4

4

•o

— o

§
rH

h

CD

O:

COrH

77S

•* u>

g
|

t^
<M

CO
CM

e

4

)

c

>

H

C

1

O

CO

O
O

1C

5

05
rH OO5
1 1 rH
•^ IO

h

>o

CM

•*
(N

H

s

1C CM l~—

CO
CM

PH

~

1 1 rH

s

u

^

£

-.-

00

2

,-H

I

~

rH 00 rH

1 1 <M

a

CM

Tf IO

^

i

£J

2;

0

Cl

GO
C.OCM

TjH 1C
0

77S

IO»O

: oo

§s

co

rH

H

3

a

ec

- *

"S "d
^ ».2

^"5 fe

•*
co

CM

O

O

oo

; -GO

.2^3 P-
III

Z'Z*

tgsi

t^

CO

»o
•*

co

•• CM

•^

.2^

•^ 9-3

Js

PH (

N '

« •

*j

Q

c i

*« C

COSOrHCMCOTflOCI

^

tJH

03 c« 3

H Jfe

&

>H

—

~

r.

»

r-

r.

r- c

7s OS OS OS OS OS OS OS C

122 BULLETIN No. 251 [May,

for fifteen years would extend from April 18 to May 21 or thirty-four
days.

The accompanying diagram shows graphically the length of the
flowering period for each year and the relation, as to length of period,
of one year with . another. Only in the year 1906 did all of the 106
varieties bloom. For the other years the numbers of varieties bloom-
ing ranged from 76 in 1901, to 103 in 1904, 1911, and 1912, 104 in
1905, and 105 in 1910. The average number of varieties flowering
each year falls between 96 and 97.

EARLY BLOOMING VARIETIES

Dates of opening of first flowers on earliest blooming varieties
ranged from April 2 in 1910 to May 5 in 1904, a difference of thirty-
four days. The average date for the sixteen years was April 24. In
nine of the sixteen years this date of first open flower was later than
the average and in the other seven years it was earlier.

Numbers of varieties opening flowers on the first day of bloom, in
the different years, vary between 1 and 46. The average was between
8 and 9 (8.68).

In 1907 Early Ripe opened the first flower April 20 and was the
only variety recorded as having open flowers on that date ; it was
followed by 7 varieties on April 21, 15 on April 22, 2 on April 23, 15
on April 24, 2 on April 25, 22 on April 26, 7 on April 27, 1 on April 29,
2 on April 30, 4 on May 1, and 4 on May 2, so that for this year the
varieties, in the opening of first flowers, were spread over thirteen
days and opened flowers on twelve of these days. Early Ripe is a
fairly consistent early bloomer ; it appeared on the first day of bloom
in eight of the fifteen years for which it has record; it appeared
on the second day in 1912, in which year it was preceded by two
varieties ; it opened flowers on the third day in each of four years —
in 1901, with 17 varieties one day in advance of it; in 1902, with 15
varieties one day in advance; in 1903, with 7 varieties leading; and
in 1906, with 8 varieties in advance. In 1911, it opened its first
flower on May 2 in company with 17 other varieties, while 44
varieties were from one to four days in advance of it. Most varieties
exhibited wider fluctuations in regard to relative position on first
bloom than did Early Ripe.

The only variety that showed greater tendency to constancy of
position in relation to flowering of other varieties was the Russian
variety, Zuzoff, which appeared in the lists of first blooming varieties
in eleven of its sixteen years of record. For the other five years this
variety appeared on the second day of bloom in each year and had
from 1 to 13 other varieties one day in advance in opening blossoms.

Oldenburg, which is one of 7 varieties having May 1 as the earliest
date of open flowers in 1901, one of 9 having April 28 as the earliest

1924} BLOOMING PERIODS OF APPLES 123

bloom date in 1902 and appearing in similar groups in three other
years, was one of 22 varieties to open the fii*st flowers on April 24 in
1903. This was the seventh day of bloom and 35 varieties were
recorded as having open flowers on one or other of the six preceding
days. In 1905 Oldenburg did not open flowers until the eighth day
of bloom. It was one of 15 varieties opening the first flowers on April
26 while there were 5 varieties with open flowers on April 19, and
65 with open flowers on dates between April 19 and April 26. For
the year 1905 there were 104 varieties blooming and of these 70, or
67.3 percent, had earlier dates of first open flowers than did Olden-
burg. In seven other years varying numbers of varieties had open
flowers earlier by one to five days than did Oldenburg.

Other varieties show the same erratic shifting of relative flowering
dates from year to year ; some with more radical changes than others,
but all showing that varietal flowering dates are in no sense fixed, but
vary within wide limits.

Causes for varietal fluctuations in relative dates of first open
flowers must be looked for in individual differences in the vigor of
the trees. All trees are subjected to the same wind, moisture, and
temperature conditions, and presumably the soil is uniform, but all
do not behave in the same manner with reference to any of their
functions; one tree, in a given year, makes greater growth than
another, has more and better foliage, and exceeds its neighbor in fruit
production; in another year the performance of the two may be
exactly reversed in all these particulars.

In the same way response of flower buds to the stimulus of advanc-
ing spring may be quick and vigorous or tardy and weak depending
upon the condition of the individual.

LATE BLOOMING VARIETIES

While some varieties have a general tendency to bloom early,
others exhibit the opposite tendency, but no variety in either group
holds a particular position in relation to other varieties with reference
to constancy. The period over which the opening of first flowers
extends is extremely variable from year to year and its length is gov-
erned largely by temperatures. A continued warm wave will group
all varieties within a short period, as was the case in 1915, when the
period from the opening of first flowers on the earliest blooming
variety to the opening of flowers on the latest blooming variety was
only five days. Intervention of lower temperatures, on the other
hand, will greatly extend the period. Thus in 1905 there were seven-
teen days between opening of first flowers on earliest and latest varie-
ties. For the other fourteen years this period varies between eight
and fourteen days, with an average for all years of a fraction over
ten days.

124 BULLETIN No. 251 [May,

The varieties that most consistently held the position of latest
bloomers were Wythe and the Russian variety, Repka Malenka.
Wythe occupied the position of latest bloomer, alone in 1905 and
1908, in company with Repka Malenka in 1911, in company with
Repka Malenka and Hall's No. 3 in 1910, in company with 3 other
varieties in 1903 and 1913, and with 4 others in 1906 and 1915. Thus
of the twelve years in which Wythe bloomed it had the latest date
of first open flowers in eight years. For the other four years Wythe,
in 1902 had the same date of first bloom as 15 other varieties ; it was
earlier by one day than 4 varieties, and earlier by three days than
Repka Malenka. In 1904 Wythe stood alone with first bloom on May
11, but was two days earlier than Sandy Glass. In 1912 Wythe had
first bloom on May 3, in company with 8 other varieties, and was one
day in advance of Repka Malenka, Garfield, and Hall's No. 3. For
1916 Wythe had the same date of first bloom as 11 other varieties,
bloomed one day earlier than 13 other varieties and two days earlier
than Repka Malenka.

Repka Malenka had the latest date of first open flower in ten of
the fourteen years in which it flowered ; it occupied this position
alone in 1902 and 1916, and was accompanied by Wythe or from
1 to 4 other varieties in the other eight years. In 1905 Repka Malenka
was four days, and in 1908 two days earlier than Wythe ; in 1904
it was four days earlier, and in 1914 three days earlier than the latest
flowering varieties.

Other varieties that have had place among latest bloomers are
Hall's No. 3 in 1906, 1909, 1910, 1912, 1913, and 1915; Rome in
1906, 1909, 1913, and 1915 ; Grimes in 1903 ; Willow in 1913 ; Hunts-
man in 1914 ; Tolman in 1903 ; and about a dozen other little-known
varieties, mostly Russian.

As before stated, no variety has proved constant in its flowering
position as related to other varieties, but perhaps there is less shifting
of dates and positions among the late flowering than among the very
early flowering varieties.

BLOOMING PERIOD OF 1910

For the year 1910 the seasonal abnormality of bloom was so great
that attending circumstances deserve mention in some detail. As
stated, the first open flowers appeared April 2 and petals were fall-
ing from the latest blooming variety on April 19, a period of eighteen
days. A period of the same length was recorded for 1909, one a day
shorter in 1904, and another a day longer in 1905. Nine of the re-
maining twelve years had periods from two to eight days shorter and
for three years the periods were three and four days longer.

The varieties commencing bloom on April 2 were Early Ripe and
Zuzoff; these were followed April 3 by open flowers on 9 varieties;

.in

1924] BLOOMING PERIODS OF APPLES 125

namely, Borovinka, Borsdorf, Hargrove, Isham,"0simoe, Ostrakoff,
Red Anis, Red Stripe, and Yellow Siberian Crab. First bloom
appeared on 11 varieties April 4, on 20 varieties April 5, on 8 varieties
April 6, on 16 varieties April 7, on 12 varieties April 8, on 19 varieties
April 9, on Garfield, Grandmother, Ox Sweet, Roe's Tolman, and
Rome on April 11, and on Hall's No. 3, Repka Malenka, and Wythe
on April 12.

The varieties flowering number 105 covering the full list with the
exception of Bogdanoff. Of the two varieties opening flowers April 2,
Zuzoff completed its period, that is the petals were falling, on April 9,
and April 10 is the date recorded for falling petals of Early Ripe.
The last variety to reach the stage of falling petals was Repka Malenka
on April 19, and of the two varieties having the same beginning date,
April 12, Wythe reached the end of its flowering period April 17,
and Hall's No. 3 on April 18.

Individual periods for the 105 varieties that bloomed ranged in
length from four to nine days, divided as follows:

Periods of

— four days 7 varieties

— five days 11 varieties

— six days 29 varieties

— seven days 23 varieties

— eight days 22 varieties

— nine days 13 varieties

These individual flowering periods were not affected by the
abnormal earliness of bloom ; they ranged much the same as in normal
years. It appears that the whole flowering period was simply moved
forward until the end of bloom, for the latest flowering variety and
the earliest opening of flowers in any other year overlap by only
one day.

The earliness of bloom is directly attributable to the abnormal
meteorological conditions that prevailed during March. This month
was dry ; rain fell on only four days, the 18th, 19th, 26th, and 30th,
with an aggregate precipitation of only .38 inch. The first two days
of the month were recorded as cloudy ; of the days following, eighteen
were partly cloudy and eleven clear. The mean daily temperature
for the month was 52.3°, 12° above normal; the mean minimum,
37.7° ; and the mean maximum, 63°. The last frost of the month
occurred on the morning of March 21, and there were early morning
frosts on thirteen of the preceding days.

For the last fifteen days of the month the mean temperature was
58.4° ; the mean maximum, 73.9° ; and the mean minimum, 44.7°.
The absolute maximum was 85° on March 24. There were four days

126 BULLETIN No. 251 [May,

on which the maximum was above 80° and ten days on which it was
above 70°.

For the eighteen-day period of bloom, April 2 to 19, rain ag-
gregating .82 inch fell on eight' days ; the last three days were cloudy,
eleven days were partly cloudy, and four days were clear. The mean
daily temperature was 53.4° ; the mean maximum, 65.6° ; and the
mean minimum, 42.1°.

The mean daily temperature for the flowering period was from
1.4° to 5.3° lower in three years, namely 1904, 1907, and 1908, and
from 2.2° to 13.7° higher in the twelve other years.

COMPARISON OF THE PERIODS FOR THE YEARS 1904 AND 1910

Having given the temperatures for the month preceding and for
the period of bloom of the year 1910 in which blooming was abnormally
early, it will be of interest to make comparison with the blooming
period for 1904, in which year the blooming period was the latest of
those recorded. In 1904 there were 103 varieties blooming, two less
than in 1910. '

Three varieties flowered in 1910 that did not flower in 1904, and
one variety flowered in 1904 that did not flower in 1910. Eliminating
these four varieties limits the comparison of the two years to 102
varieties that have records for both years. The full flowering periods
for the two years differ by but one day ; seventeen days in 1904, and
eighteen days in 1910, but the beginning date in 1910 was thirty-four
days in advance of the beginning date in 1904. The two dates were
April 2 in 1910 and May 5 in 1904. As the periods differ in length
by one day the ending dates, April 19 in 1910 and May 21 in 1904,
are separated by thirty-three days.

Flowering periods for individual varieties tended to concentration
about the six-, seven-, and eight-day periods in both years; in 1904,
77 or 75.49 percent of the varieties fell under these three periods and in
1910, 71 or 69.60 percent of the varieties were recorded as in these
periods.

The day-groups in 1904 numbered eight, with a range of four to
fourteen days ; in 1910 there were six day-groups with a range of four
to nine days. In 1904 there were two varieties having ten-day periods
and one with a fourteen-day period. Distribution of varieties into
periods differing in length by one day was as follows:

1924] BLOOMING PERIODS OF APPLES 127

Varieties having each a period of

1904 1910

— four days 3 7

— five days 6 11

— six days 16 27

— seven days 28 22

— eight days 33 22

— nine days 13 13

— ten days 2 0

— eleven days 0 0

— twelve days 0 0

— thirteen days 0 0

— fourteen days 1 0

102 102

The average period was 7.34 days for 1904 and 6.78 days for 1910 ;
longer by .56 day in the year when the blooming period occurred
thirty-four calendar days later than in the year of early bloom. It
would be reasonable to expect that in contrasting the blooming periods
of two years, one of which was more than a month later in the season
than the other, the periods of individual varieties would be shorter
and concentrated under fewer day-periods in that year in which the
blooming time was latest in season. This is expected because the more
powerful action of the sun and the presumed greater aggregate of
heat units in the advanced season should so stimulate plant functions
that blooming of varieties would proceed rapidly and result in con-
centration within narrow time limits of short individual periods. In
this case, however, the reverse was true. Not only were the full flower-
ing periods, altho separated by more than a month, of nearly equal
length, but individual periods had a shorter average length in the early
year and there were twice as many varieties having four- and five-day
periods as in the later year.

Temperatures preceding and during the blooming period of 1904
may now be examined, considering first a period of thirty-two days
from April 3 to May 4, which equals the period March 1 to April 2
considered in relation to the bloom in 1910. For this period the maxi-
mum temperature was 80° ; minimum, 22° ; and mean, 46.60°. Dif-
ferences between the temperatures for the two years are most readily
compared by placing them in parallel columns as below.

1904 Maximum Minimum Mean

April 3-May 4 .Absolute 80° Absolute 22° 46.60°

Mean 57.97° Mean 35.84°
1910

March 1-April 1 Absolute 85° Absolute 17° 52.30°

Mean 63° Mean 37.7°

The absolute maximum for the early period of 1910 was 5° higher
than for the later period of 1904 and this difference holds approxi-
mately for the mean maxima of the two periods.

The absolute minimum for 1910 was 5° lower than for 1904, but
the average of minimum temperatures for the periods under consid-

128 BULLETIN No. 251 [May,

eration (for the two years) was nearly 2° higher for 1910 than for
1904. Mean temperatures for the periods were 46.60° for 1904, and
52.30° for 1910 ; the mean for the year of early bloom 5.7° higher than
for the year of late bloom.

Comparing actual blooming periods — April 2 to 19 in 1910, and
May 5 to 21 in 1904 — it is found that higher temperatures prevailed
in 1904 than in 1910. The maximum in 1904 was 84°, in 1910, 80°,
with an average maximum in 1904 of 70.05° and in 1910 of 65.6° ; the
1904 average was 4.45° higher than in 1910. Minimum temperatures
were 41° for 1904 and 30° for 1910; the average minimum, 48° for
1904 and 42.1° in 1910; the 1904 average was nearly 6° higher than
the average of 1910. Mean daily temperatures were 58.91° for 1904
and 53.4° for 1910; that for the year of late bloom was 5.51° higher
than that for the year of early bloom.

For the thirty-two days preceding bloom in 1904, temperatures of
32° or lower were recorded on twelve days ; for the similar period in
1910, on ten days. During the blooming period in 1904 there were
no frosts, but in 1910 frosts occurred on five days.

From this examination of temperatures, it appears that the heat
stimulus for the period of thirty-two days preceding the first opening
of flowers was somewhat greater from March 1 to April 1 in 1910 than
it was from April 3 to May 4 in 1904, and that for the periods of
bloom there were higher temperatures for the May period of 1904
than for the April period of 1910.

The chief reason for the unseasonable advance of flower buds in
1910 appears to rest in the high temperatures that prevailed during
the period of thirty-two days just preceding £he opening of flowers.
The temperatures thruout the period of blooming continued suffi-
ciently high to concentrate varietal periods into six groups of four to
nine days, with a shorter average individual period than is recorded
for 1904, when the bloom came a full month later in the season.

To what extent differences in moisture supplemented the tempera-
ture differences between the two seasons compared is not known, but it
seems probable from differences in precipitation that the influence
may have been important. For the thirty-two-day periods preceding
bloom in the two years the rainfall and cloudiness are recorded as

follows :

Bainfall No. of days on Clear Partly cloudy Cloudy

(inches') which rain fell days days days

1904 3.44 7 10 18 4

1910 38 4 12 18 2

Rainfall for the 1904 period was nine times that of the 1910 period.
Clear days were two less and cloudy days two more than in 1910. The
heaviest rainfall of 1910 was less than the least fall in 1904 and the
heaviest single rain (1.61 inches) in 1904 was more than four times the
total for the period in 1910.

BLOOMING PEFJODS OF APPLES 129

For the seventeen-day duration of bloom in 1904, the rainfall was
.94 inch, distributed over nine days, two of which record only a trace.
In 1910 the rainfall for the eighteen-day period of bloom was .82 inch,
distributed over ten days, and again two of the days record a trace
only. Five of the seventeen days in 1904 and four of the eighteen days
in 1910 were clear. Eleven days of the periods in each of the two years
were partly cloudy; one in 1904 and three in 1910 were entirely
cloudy.

After blooming began, varieties responded much more rapidly in
1904 than they did in 1910. This is shown by the fact that only two
varieties opened flowers on the earliest date of bloom in 1910 and at
the end of the fifth day from the beginning of bloom only 50, or 47.6
percent of all varieties had open flowers ; whereas in 1904 there were
thirteen varieties that opened flowers on the first day of bloom and at
the end of the fifth day 99, or 96.11 percent were recorded as having
open flowers.

Comparing flowering periods for individual varieties, it appears
that 23 varieties had periods of equal length in the two years, while
for 79 varieties the periods were either longer or shorter in one year
than in the other. Thus 19 varieties had periods one day longer in
1904 than in 1910, and 13 varieties had periods one day longer in 1910
than in 1904. There were 13 varieties with periods two days longer
in 1904 than in 1910, and 12 varieties with periods two days longer in
1910 than in 1904 ; 13 varieties whose periods were three days longer
in 1904 than in 1910, and 3 whose periods were three days longer in
1910 than in 1904 ; 3 varieties with periods four days longer in 1904
than in 1910, and only one with a period four days longer in 1910 than
in 1904. Two other varieties exhibited still wider differences. Grand-
mother, which had a five-day period in 1910, had a ten-day period in
1904, a difference of five days, or a period twice as long in the year
of late bloom as in the year of early bloom. The extreme of difference
is seen in periods of Repka Malenka, which has a record of eight days
in 1910 and of fourteen days in 1904, a period six days longer in the
year of late bloom than in the year of abnormally early bloom.

Fluctuations in length of flowering periods from year to year are
the rule rather than the exception and further illustrations will be
given elsewhere. But the differences between periods in these two
years, and the fact that for 50 varieties the periods were from one to
six days longer in 1904 than in 1910, and that 29 other varieties had
periods from one to four days longer in 1910 than in 1904, suggests
the operation of influences aside from temperature and moisture and
about which nothing definite is known.

130 BULLETIN No. 251 [May,

VARIETAL FLOWERING PERIODS

For the different years the average varietal flowering period for
all varieties blooming in the particular year, ranged from five to ten
days. This average was five days in 1901 and 1915, six days in 1908
and 1912, seven days in each of six years, eight days in each of three
years, nine days in 1903 and 1914, and ten days in 1907.

The extremes in length of blooming periods for individual varieties
may be near together in one year and widely separated in another
year ; there is great variation in this. Thus in 1907, a year in which
only 82 varieties bloomed, the minimum period of three days was rep-
resented by 2 varieties, the maximum period of seventeen days by 1
variety, and each number of days falling between three and seventeen
was represented by 1 to 13 varieties ; this largest number, 13, falling
to the twelve-day period. The total number of periods, each differing
from its neighbor by one day, was fifteen. At the other extreme, the
year 1915 had the record of a concentration of 100 varieties into three
blooming periods ; 7 varieties had each a four-day period, 57 varieties
had five-day periods, and 36 varieties had six-day periods. Next to
1915 in number of periods differing by one day, is the year 1902 with
four periods ranging from five to eight days in length. The 87
varieties were distributed as follows: with five-day periods, 5
varieties ; with six-day periods, 34 varieties ; with seven-day periods,
31 varieties, and with eight-day periods, 17 varieties. Each of the
years 1901 and 1912 concentrates the flowering record into five groups
ranging from four to eight days. Other years range thru six, seven,
eight, nine, ten, and eleven groups each, and then follow the maximum
of fifteen groups in 1907, as already mentioned.

Concentration into few groups, or distribution into many groups,
follows the full length of flowering period ; that is, from the opening
of the first flower on the earliest blooming variety to the fall of the
petals on the latest blooming variety, with a regularity which, altho
not perfect, approximates the expectation. The shortest blooming
period was ten days in 1915, and in that year appeared the maximum
of concentration : three groups of four, five, and six days, respectively.
The longest blooming period of twenty-two days was recorded for the
year 1907, in which year occurred the maximum distribution into
fifteen groups with periods three to seventeen days in length.

Next below the maximum in length of flowering period come the
two years 1903 and 1908 each with a period of twenty-one days. In
the earlier year the 99 varieties flowering fall into ten groups with
varietal periods ranging from five to fourteen days; in 1908 the 97
varieties fall into eleven groups with varietal periods ranging from
three to sixteen days, but with no representatives for the nine-, eleven-,

1924]

BLOOMING PERIODS OF APPLES

131

and fourteen-day groups ; 90 of the varieties were in the five groups
with three to seven-day periods; the remaining seven were scattered
in six groups of greater length.

The full flowering period in 1905 was nineteen days, that of 1909,
eighteen days ; in both years the number of groups was eleven, but in
1909 the longest period was one day longer than the longest period in
1905.

Distribution of varieties into groups according to length of bloom-
ing period is shown for each year in the accompanying table, together
with the number of varieties blooming and the average varietal period
for each year.

TABLE 2. — DISTRIBUTION OF VARIETAL BLOOMING PERIODS IN EACH YEAR

Year

Number
of
varieties
blooming

Length of blooming period in days

Average
varietal
period,
diiys

3

4

5

6

7

8

9

10 | 11

12

13

14

15

16

17

1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916

76
87
99
103
104
106
82
97
100
105
103
103
102
84
100
94

39

19
5
7
6
5
6
1
34
20
11
11
42
15
4
57
23

11
34
10
17
10
32
8
32
21
29
18
32
25
6
36
8

6
31

5
28
11
39
5
9
12
23
12
11
12
1

1
17
26
33
25
17
10
1
25
22
34
9
11
9

4.83
6.69
9.20
7.30
8.23
7.00
9.93
5.89
7.02
6.77
7.35
5.70
7.58
9.03
5.29
7.02

4
13
21
10
6

ii

13
17

12
2
21
1
12
1
3

13

16

2

4
1

"3

3
1

3
1
11

"2

3

1

2
2
1

1
13

3
7
4
9
1

13

1
1

4
1

4

i

3
2

1
1

1

7

12

29

21
27

5
4

3

1

7
4

16

19

23

1

AMOUNT OF BLOOM

While record of the amount of bloom has been kept under the five
gradations — none, scant, moderate, full, and very full — as elsewhere
explained, it is necessary to consider, from a practical standpoint, only
three divisions ; namely, no bloom, amount too small to suggest a crop
of fruit, and sufficient to promise a crop. The amount of bloom is not
open to exact determination and the record, at best, is only an estimate ;
but effort ha« been made to keep the basis of estimate as uniform as
possible. It is believed that the record will be helpful in the summa-
tion of the characteristics of varieties, in segregating those with well-
defined tendencies towards insufficient bloom from those having the
opposite tendency, and in advancing studies that may be undertaken
into the causes for blooming tendencies exhibited.

Observation of the blooming of apple trees year after year conveys
a very definite impression of wide seasonal differences in behavior of

132 BULLETIN No. 251 [May,

varieties and individuals, of very unequal response to exterior condi-
tions observed and assumed to exert influence either as stimuli to in-
creased activity or as agents operating to retard or diminish plant
processes, and of the fact that factors governing performance are
complex and dependent upon physiological changes that are difficult
to understand or interpret rightly.

Amount of bloom is not determined at blooming time, as the buds
are formed the preceding summer; they make small advance in the
fall and may even make slight growth during the winter. They push
and make rapid or slow development in the spring, according to the
climatic conditions of the season, but whether the amount is large or
small in a large measure must be controlled by conditions and sur-
roundings of the trees during the formative period, nine or ten months
preceding the opening of flowers in spring.

Classified into the three divisions — no bloom, not enough for a
crop, and sufficient for a crop of fruit — the 1,696 records for the 106
varieties for sixteen years divide as follows : no bloom, 151 ; not
enough to make a crop, 485 ; and enough to promise a crop, 1,060.
Deficiency records constitute 371/2 percent of the total, leaving 62!/o
percent to represent cases of sufficiency or excess of bloom.

The 151 records of failure to bloom are distributed in numbers
ranging from 1 to 30 thruout fifteen years ; only in the year 1906 did
all varieties produce flowers. The larger number of failures, 30,
occurred in 1901 ; in 1907 the number was 24; in 1914, 22; in 1916,
12; with numbers less than 10 in the other years.

Records of scant or insufficient bloom are much more numerous
than those of no bloom, constituting, as they do, 28.6 percent of all
records. Numbers of varieties appearing in this class, in the different
years, range from 1 to 59 and are as irregular in distribution as are
numbers in the class having no bloom.

In thirteen of the sixteen years, numbers of varieties in the scant
bloom column are larger than in the no bloom column, but it does not
follow that high numbers in the scant bloom column are always ac-
companied by low numbers in the no bloom column ; thus in 1901 with
the maximum of 30 in the no bloom column there are 33 in the scant
Moom column, and in 1907 the respective numbers are 24 and 41. On
the other hand, the two highest numbers of varieties in the scant bloom
column are associated with low numbers in the no bloom column, as
59 scant bloom and 3 no bloom in 1911 ; 51 scant bloom and 2 no
bloom in 1905. For the three years 1913-1915, the reverse is true;

1913 has 1 in the scant bloom column to 4 in the no bloom column ; for

1914 the numbers are 14 and 22, and for 1915, 3 and 6 respectively.
The most satisfactory index of performance as to amount of bloom

is found in the distribution of varieties between the two groups, in-
sufficient or no bloom, and bloom sufficient for a crop.

1924} BLOOMING PERIODS OF APPLES 133

Examining the record, it appears that 1913 was the year of maxi-
mum performance, for 101, or 95.28 percent, of the 106 varieties car-
ried sufficient bloom. Next to this is 1915 with 97, or 91.51 percent,
of the varieties having satisfactory production. At the other extreme
are the years 1902 and 1907, each having only 41, or 38.68 percent, of
the varieties in the group having sufficient bloom. The twelve other
years have numbers of varieties in this same group ranging from 43 to
91, with percentages from 40.57 to 85.85.

It is, perhaps, worthy of note that only ninety-two records appear
of varieties having exceptionally heavy bloom; these are distributed
by years as follows : 30 varieties in 1914, 22 in 1915, 15 in 1902, 11 in
1904, 4 in each of the years 1901 and 1916, and 3 in each of the years
1903 and 1912. For the seven consecutive years 1905 to 1911, and for
the year 1913 — eight years in all — there are no varieties represented in
the enumeration of cases of very heavy bloom. The varieties that do
have place among those having very full bloom in one or more years
number 58 ; of these 32 appear in the list but once, 21 appear twice,
3 appear three times, 1 four, and another five times. There are three
cases in which the years of extra heavy bloom were consecutive and all
for the years 1914 and 1915; the varieties are Pointed Pipka, Red
Aport and Borsdorfer, the latter, eleven years before was credited with
a third year of heavy bloom. There are also twelve cases involving 9
varieties in which years of very heavy bloom were separated by only
one year, and that year had full bloom in one case, moderate bloom in
seven cases, scant bloom in three cases, and none in one case.

The relation of very heavy bloom to bloom in preceding and
following years has been examined and from the results it may be
stated that in four cases where heavy bloom occurred in the first year
of record, the performance of the preceding year was unknown; of
the eighty-eight remaining records of the next preceding years, sixty-
two, or 70.45 percent, had either full or moderate bloom (enough to
promise a crop), while twenty-six, or 29.55 percent, recorded scant
bloom or none. Following heavy bloom there were four years in which
heavy bloom occurred in the last year of record with no record for the
following year. The eighty-eight records of following years divide as
follows: sufficient bloom for a crop in fifty-seven, or 64.77 percent,
and insufficient or no bloom in thirty-one, or 35.23 percent.

For the twenty-one varieties, each appearing twice in the list of
those recording heavy bloom, there are two cases in which the years of
heavy bloom were consecutive, five cases in which there was alterna-
tion, with sufficient bloom in the intervening year, and fourteen cases
in which the years of very heavy bloom were separated by eight to
twelve years.

134 BULLETIN No. 251 [May,

The three varieties each recorded as having heavy bloom in three
years were Borsdorfer, Hibernal, and Willow. Borsdorfer had the
highest record of any variety, for in not one of the sixteen years was
there a deficiency in bloom ; the record of this variety began with full
bloom in 1901, extra heavy bloom in 1902 followed by three years of
full bloom, a year of moderate bloom, then full bloom, again followed
by a year of moderate bloom, then five consecutive years of full bloom
followed by very heavy bloom in 1914 and 1915, ending with full bloom
in 1916. Hibernal, while having very full bloom in 1902, 1904, and
1915, and sufficient for a crop in four other years, had no bloom in
1901, 1907, and 1914, and bloom insufficient for a crop in 1903, 1905,
1912, and 1916. A summary for this variety shows nine years of
sufficient bloom and seven years of deficient bloom.

Willow began with very full bloom in 1901 and duplicated the per-
formance in 1903, with a year between of scant bloom; it had full
bloom in 1904 and of the following ten years, six had each enough for a
crop and four less than enough, then came a year df very full bloom in
1915 followed by a deficiency in 1916 ; or, for the sixteen years, ten
had enough bloom and six had deficiencies altho in no year was there
entire absence of bloom.

The variety recorded as having very full bloom in four years was
Oldenburg; the years of very full bloom were 1902, 1904, 1914, and
1916 ; between each of these pairs was a year of moderate bloom. The
record began with scant bloom in 1901, then for the years 1905 to
1913 inclusive were six years of bloom sufficient for a crop, two years
with insufficient bloom, and one year with no bloom — a record for the
sixteen years of twelve years in which the bloom was good and four
in which it was poor.

Borovinka reached the maximum with a record of five years of very
full bloom. These years were 1902 and 1904, preceded, separated, and
followed by years of scant bloom; 1912, preceded by a year of scant
bloom and followed by a year of moderate bloom ; and 1914 and 1916
with a year of no bloom between. The years 1906 to 1911 were repre-
sented by four years of bloom sufficient for a crop, one with insufficient
and one with no bloom. Thus for the sixteen years, ten had sufficient
bloom and six were deficient in bloom.

There are wide differences in varieties with regard to their bloom-
ing tendencies. Some varieties are inclined to produce sufficient bloom
nearly every year while others rarely attain satisfactory production.
Borsdorfer, as already mentioned, stood alone writh record of sufficient
bloom in each of the sixteen years. Then there were five varieties —
namely, Arabian, Berry, Osimoe, Rome, and Zuzoff — each with de-
ficiency in only one year, and six varieties — Arkansas, Borsdorf,
Grandmother, Hargrove, Twenty Ounce, and Yellow Siberian Crab —

1924} BLOOMING PERIODS OF APPLES 135

each with two of the sixteen years recorded as deficient in bloom. In
another group of eleven varieties including Fameuse, Golden Ball,
Roe 's Tolman, Wolf River, and Wythe, each variety divided equally
between years of sufficient bloom and years of deficiency. In still
another group of twenty-two varieties each had from nine to fourteen
years of deficiency, with two to seven years of sufficient bloom ; among
these were Huntsman, Jefferis, and Yellow Sweet, each with eleven
lean years and five full years ; Blackwood, Bogdanoff, Red Aport, and
three others stood ten poor to six good. Most persistently unproduc-
tive of bloom were Peach and Rambank Gasser, each with record of
fourteen years of deficient or no bloom and only two years with enough
bloom to promise a crop.

It is possible to pick out, here and there in the record, instances of
the alternation of full or heavy with light or no bloom, but occurrence
of these instances is extremely irregular and much less common than
the occurrence of three to five and even six consecutive years charac-
terized by the same degree of bloom, either moderate or full or scant.

In this fact is seen evidence of the inequality of response by dif-
ferent varieties to the same attendant conditions. Thus for the four
years 1909 to 1912, Borsdorf and Borsdorfer bloomed full in each of
the years, while Fameuse had scant bloom in the first three years and
moderate bloom in the last, Garden scant bloom in the first three years
and moderate bloom in the last, Jefferis and Kruder scant bloom in all
four years, and so they vary thruout the list. Each variety appears to
follow a course of its own. Individuality stands out strongly and there
appears no single agency nor group of agencies that operates on more
than very restricted lists of varieties for limited periods, to govern per-
formance in flower production.

Distribution of varieties in regard to amount of bloom in the dif-
ferent years is given in Table 3. Arranging the same records to bring
the varieties into two groups on the basis of deficient or no bloom, and
sufficient for a crop, and adding the percentages in each group for each
year, the numbers of varieties distribute as shown in Table 4.

For the sixteen years the average percentage of varieties having
bloom sufficient for a crop was 62.5 and there was equal division of the
years, eight falling below the average and eight ranging above, but
with no regular order nor alternation. The years falling below average
were 1901, 1902, 1905, 1907, 1908, 1909, 1911, and 1912 ; those above
average were 1903, 1904, 1906, 1910, 1913, 1914, 1915, and 1916.

Effort has been made to trace a relationship between the amount
of bloom in any given year and conditions of temperature and mois-
ture prevailing during the period of bud formation — that is to say,
during the months of June, July, and August of the preceding year —

136

BULLETIN No. 251

[May,

_,

1C

CD

N

-«

«c

00

1

CO

O5

CD

M

00

(N

0

OS

1-H

1C

(N

>0

CO

CO

5

00

^

O

i-H

£

1— I

OS
CM

1-H

0

OS

* - 5- 3 °

OS

co

9

co

01

ro

i-H
OS

co

OS
1C

co

<N

00

O

O
OS

iH

co

i-H

(M

IS

O

os

o

OS

CD

5

co
co

CO

c

00

o

OS
i-H

os

CO

co

co

CM

0

o

•^

.

<*

o

2

CM

"*

CO

OS
i-H

0

c5

CD

CD

co

~

c

OS

(N

i-H

OS
CM

gj

0

T-H

o
o

CO

c^

OS
CM

i-H

1C

-

co

o

OS

-

<N

os

o

CN

c«

c

Os

OS

CO

„

CM

r^

i-H

o
os

0

co

CO
CO

CO
i-H

co

CM

*

a

. v

£

o

_o

1s

i O Q}

s

o

S3

Q

03

c

9

0 J

o •-

I^S

H c<i

C-H
CO

>

T*

R

i
"9

1

CO

o

o co

(N

O
CO Is-

1,

1

f
00

OS

a
t

os

rn

CO OS

co

00
t-cd
CO

3

co

(N

1C •>

00

1-H (M

4
t

os

1C"*

1C
1C

H

z
a

3

-H

OS
1-H

<M -^

'-OGO
>c

•*T-H"

•I

3
Q

a

O

1-H

os

H

o

0006

co id

1C
CD

Z
<

*

5

OS

o

os

o

sS

co

CD 00
10 01
1C

q

H

a

00

o

OS
i-H

49

(M «O

co

|

00

•HI CO
CO

H

3

b

J

c

OS
i-H

o>

p

(N

ic co

CD

"S

5

'S

00

1-H CO

co

Q

3

OS

<a
o

fe

Tt< CD

^ oi

(N

OQ

CO

CM CO

i

3

3

o

OS

0

cop

1C

co|

1C

H

2
->

8

Tf

o

OS

1C

ic*-;

I-H

1C

^H 00

os >c

00

a

-i

i

co
o

OS

CO c^j

CO

Ol OS

CD

3

4
>

•i

(N

o

OS

(M

CD

00

,-H CD

"* 00
CO

3
;

H
H

i-H

o

OS
i-H

CO

co •*
•c

CO "?

|

-1

Q

H

i.
i

i

4

3

cu

No. of varieties
Percentage

No. of varieties
Percentage

19S4] BLOOMING PERIODS OF APPLES 137

but the irregularities found were such as to render doubtful the exist-
ence of any definite relation.

The summer months of 1912, preceding the year with the maximum
percentage (95.28) of varieties having sufficient bloom (1913), had a
mean monthly temperature of 71° and rainfall of 7.63 inches. Next
to 1913 was 1915, with 91.51 percent of the varieties having sufficient
bloom. For the summer of the preceding year, the mean monthly tem-
perature was 78.2°, the highest of any year, and the rainfall 6.49 inches.

Record of these years suggests the possibility of high percentages
of varieties having abundant bloom in years following high summer
temperatures and low rainfall; but other years with preceding sum-
mer temperatures almost as high and with rainfall higher, but still
below normal, do not exhibit the same high percentages of varieties
with ample bloom. Thus 1911 with only 41.51 percent of the varieties
having sufficient bloom, had, in the preceding summer, a mean monthly
temperature of 71.8° and a rainfall of 8.37 inches ; 1907, with 38.68
percent of the varieties with bloom enough for a crop, had, for the
preceding summer period, 72.5° as the mean monthly temperature and
a rainfall of 9.81 inches.

Perhaps the irregularities of the relation between the amount of
bloom and the weather conditions for the preceding summers are
rendered most evident by grouping and comparing the highest and
lowest percentages of varieties having sufficient bloom. Thus the
average percentage of varieties having ample bloom in the years 1904,
1913, and 1915 (90.88 percent) may represent the highest. The
average mean monthly temperature for the preceding summers was
73.3° and the average rainfall for the same periods, 8.05 inches. With
this group compare the years 1901, 1902, and 1907, the three years
having the lowest percentages of varieties with sufficient bloom, and
which have an average of 39.31 percent. Mean monthly temperature
for preceding summers averaged 73.9°, with the average rainfall 11.31
inches.

While it appears that the percentage of varieties having enough
bloom for a crop was nearly 2% times greater for the high than for
the low group, the average temperature for the summers preceding the
years of high percentages of varieties having enough bloom was less
by 0.6° than for the same periods preceding percentages in the low
years, and the rainfall for summers preceding high percentages was
less by 3.26 inches than for summers preceding low percentages.

Comparison in the same way may also be made on the basis of
highest and lowest rainfall during preceding summers. Thus the
average rainfall for the summer months of 1902, 1907, and 1915, the
years of greatest rainfall, was 18.71 inches; and the average per-
centage of varieties with sufficient bloom in the next following years,

138 BULLETIN No. 251 [May,

1903, 1908, and 1916, was 66% percent. The years 1909, 1912, and
1914, following the years of lowest rainfall (an average of 5.28 inches
for the summer months) show an average of only 58.81 percent of the
varieties with sufficient bloom. From these contrasting periods it
would appear that the highest bloom percentages follow years of high-
est rainfall.

Again comparison may be made on the basis of mean temperatures
for preceding summers. The three years having highest summer tem-
peratures were 1901, 1913, and 1914, with an average monthly mean
of 76.5° ; the average of varieties having sufficient bloom in the next
following years was 65.41 percent. Years having lowest summer tem-
peratures were 1903, 1904, and 1915, with an average monthly mean
for the period of 70.6° ; for the next succeeding years, 1904, 1905, and
1906, the average of varieties with sufficient bloom was 69.18 percent.
Here the years of high summer temperatures were followed by lower
percentages of varieties having sufficient bloom than any recorded for
years following low summer temperatures. These comparisons of high
and low temperatures make it appear that, on the basis of percentages
of trees having sufficient bloom, high percentages follow years having
slightly lower temperatures and appreciably lower rainfall than the
years preceding the low percentage years ; that on the basis of rainfall,
high percentages of varieties with sufficient bloom follow years of
highest rainfall ; and that on the basis of temperature, the high per-
centages follow the lowest temperatures.

Several other combinations have been examined, but no importance
attaches to any of them. Correlation between percentages of bloom
and temperatures and rainfall of preceding years cannot be established
for the reason that temperature and rainfall are only two of a long
array of climatic factors, all of which operate to influence bloom per-
formance.

Temperature and rainfall may be the most important agencies
affecting amount of bloom, but taken separately they cannot lead to
correct interpretation of observed results because other and possibly
equally important factors are left out. Computations from incomplete
data only serve to confuse ; they do not aid in understanding observed
phenomena and hence accomplish no good purpose.

That the wide variations in numbers of varieties producing bloom
and the equally wide variations in amount of bloom in different years,
result from definite causes is a warranted assumption, but determina-
tion of these causes is impossible in the absence of complete data, and
even with full data would be difficult because of the complex nature of
the problem. Like any other problem involving the physiological
processes of plants, the factors that influence results are very nu-
merous ; these factors interact among themselves, they are difficult to

1924} BLOOMING PERIODS OF APPLES 139

isolate; and to correctly associate any one of them with observed
phenomena is possible only thru close and prolonged study.

AMOUNT OF BLOOM AND DISTRIBUTION OF VARIETIES
INTO TIME-GROUPS

An impression, acquired from observations and unclassified notes,
that concentration of the flowering of varieties into a few time-groups
occurred in years of heavy bloom, and that wide distribution into many
groups was characteristic of years of light bloom, is not fully supported
by the tabulated records of all years. It is true that 1907, the year of
widest distribution, when there were fifteen groups ranging from three
to seventeen days, was one of two years having the minimum amount
of bloom. In 1907 only 41, or 38.68 percent, of the 106 varieties had
bloom in quantity sufficient to promise a crop ; 37 of the varieties were
recorded as having moderate bloom and 4 as having full bloom; 21
varieties had no bloom and 44 had each a few scattering flowers, not
enough, however, to suggest a crop of fruit. The year of maximum
concentration, when the varieties having bloom enough for a crop were
gathered into three groups of four-, five-, and six-day periods, was
1915, a year standing next below the maximum in amount of bloom.
In this year there were 97 varieties, or 91.51 percent of the total, with
bloom sufficient for a crop ; 47 varieties had moderate bloom, 50
bloomed full and 22 of these were very full.

These years of minimum and near-maximum bloom support the im-
pression of a relationship between amount of bloom and time distribu-
tion, but this suggested relationship is rendered doubtful by the bloom
performance of other years. For the year of maximum bloom, 1913,
when 95.28 percent of the varieties carried bloom enough for a crop,
there was distribution into eight groups ranging from four to eleven
days. In the year 1902 which, in company with the year 1907, records
the lowest percentage of bloom, only 38.68 percent, there was con-
centration into four groups ranging from five to eight days. Next
above the minimum in amount of bloom was the year 1901, with a per-
centage of 40.57 and with five groups ranging from four to eight days.
The three years 1905, 1908, and 1909, each with a distribution of
varieties in eleven groups, have percentages of varieties blooming of
50, 60.38, and 52.83 respectively. Without going further into detail
it appears that for selected pairs of years a rather definite relation
may be shown between the number of varieties blooming and the dis-
tribution of the varieties into greater or less numbers of time-groups,
but when the aggregate of records is considered no constant relation
is apparent.

140 BULLETIN No. 251 [May,

TEMPERATURE AND DISTRIBUTION OP BLOOM

Temperature for the duration of the blooming period is, without
doubt, an important factor in influencing distribution of varieties into
groups representing flowering periods of different lengths. Comparing
the record of flowering periods and distribution with the mean daily
temperatures for the respective years, it appears that in the main,
length of period and distribution fluctuate with mean daily tempera-
tures. Thus the year 1915, with the shortest recorded full flowering
period of ten days and concentration of the 100 varieties into four-,
five-, and six-day groups, had a mean daily temperature for the period
April 23 to May 2 of 67.1° ; while, the year 1907, with a flowering
period of twenty-two days and a distribution of the 82 varieties into
fifteen time-groups ranging in lertgth from three to seventeen days,
had a mean daily temperature fojPihe period April 20 to May 11 of
48.1°. Other years exhibit a somiglfchat similar relation, but there are
sufficient irregularities to indicate" that factors other than temperature
frequently operate to modify botlf length of blooming period and range
of distribution. For example, the. year 1910, with a full flowering
period of eighteen days and a dislribution of varietal periods into six
groups of four to nine days each, tifid a mean daily temperature for the
period April 2 to April 19 of 53.J^^^tiile for the year 1914, with a
full period of fifteen days and the »|ietal periods distributed in nine
groups of five to thirteen days each^he mean daily temperature from
April 27 to May 11 was 61.5°.

A further indication of the operation of factors other than tem-
perature is suggested in the fact that varietal flowering periods of
three days are recorded only in the years 1905, 1907, 1908, and 1909,
all of which had long full-flowering periods. The distribution of
varieties into the maximum numbers of groups and mean daily tem-
peratures for the respective flowering periods were as follows: 58.4°,
48.1°, 50.4°, and 55.6°.

Examination of the flowering record of all varieties arranged by
years shows that wide variation in length of varietal blooming periods
in different years is common to all, but the range varies for different
varieties. There are 8 varieties each having three days as the mini-
mum period ; for these varieties the maxima range from eight to fifteen
days. Next to this is a group of 63 varieties, each with a minimum
period of four days and here the maxima range from eight to sixteen
days. Then follows a group of 33 varieties each with a minimum of
five days and maxima ranging from nine to seventeen days. Finally,
of 2 varieties each having a six-day period as the minimum, one has a
maximum period of eleven days, the other a maximum period of fifteen
days.

Variation in length of flowering period is expected. Mean tem-
peratures for particular periods vary greatly in different years, and

BLOOMING PERIODS OF APPLES 141

flowering periods should be long when temperatures are low and short
when they are high.

While this relation between temperatures and length of flowering
periods is apparent in a general way, there are many irregularities and
such wide departures from the rule as to make it evident that impor-
tant and complex factors other than temperature and atmospheric con-
ditions at time of blooming, operate to modify the duration of flower-
ing periods. To illustrate, the flowering periods of Domine and
Fameuse for the year 1907 may be compared. The period for Domine
was six days, that for Fameuse fourteen days; the mean daily tem-
perature for the six-day period of Domine was 47.6°, that for the
fourteen-day period of Fameuse 47.4°. It does not appear reasonable
that a difference of 0.2° in mean daily temperature should be the sole
cause of prolongation of one period to 2% times the length of the other.

A similar case appears in the record of flowering periods for Rome
and Wolf River in 1903. The mean daily temperature for the Rome
period of twelve days was 53.6°, and for the five-day period of Wolf
River 53.4°. Another example is found in comparison of the records
of Shockley and Tolman, also for the year 1903. Shockley had a flower-
ing period of ten days, Tolman a flowering period of six days ; the
mean daily temperature for the Shockley period was 50.5°, that for the
Tolman period 49.6°. Here the shorter period had a mean daily tem-
perature less by 0.9° than that of the longer period. For the year 1908
these same varieties show a rather extreme difference ; Shockley had a
flowering period of six days while that of Tolman was sixteen days,
but in this case temperature may be accepted as the chief cause of
difference, for the mean daily temperature of the Shockley period was
57.6°, while that for the Tolman period was 47.1°, a difference of

101/20.

INCONSTANCY OF RELATIVE LENGTHS OP VARIETAL
BLOOMING PERIODS

With nearly all varieties there are marked irregularities in the
alternation of long and short blooming periods and, further, there are
conspicuous departures, from those relative lengths of periods of dif-
ferent varieties, that might be easily assumed to be constant ; that is
to say, where two varieties in any one year have blooming periods dif-
fering in length by several days, one perhaps twice as long as the other,
it would not be unreasonable to suppose that the varieties in question
would hold to an approximation of the difference in other or all
seasons. As a matter of fact, the relative lengths of the periods in one
season may bo reversed in a succeeding season. Such occurrence indi-
cates plainly that temperature and general atmospheric conditions,
which are commonly regarded as the chief determiners of length of
blooming periods, do not act equally on all varieties in all seasons, or

142 BULLETIN No. 251 [May,

that varieties develop within themselves qualities that render them less
susceptible to stimulation, or more resistant to adverse conditions in
one season than in another. Records for Tolman and Whitney will
serve to illustrate the fluctuations here referred to. The flowering
period for Tolman in 1903, a year having a long average flowering
period (9.2 days) for all varieties, was six days; while in 1908, when
the average period for all varieties was short (5.94 days), the period
recorded for Tolman was sixteen days. In contrast with this record
Whitney, with an average flowering period for sixteen years of 7%
days, had its flowering period lengthened to fourteen days in 1903 and
contracted to five days in 1908. Records for these two varieties for the
two years, placed side by side for ready comparison, appear as follows :

1903 1908

Blooming period for Tolman 6 days 16 days

Blooming period for Whitney 14 days 5 days

Tolman did not bloom in 1907, so that the two varieties can be
compared for only fifteen years. For the years other than the two
given above, the flowering periods of the two varieties were as follows :
they were equal in 1912, in five years Tolman had the longer periods
by one to two days, and in seven years Whitney had the longer periods
by one to five days. The trees compared were of the same age, grew
but a few rods apart and received the same care each year.

To further illustrate irregularities in flowering periods, certain
other varieties may be briefly mentioned. Ben Davis is recorded as
having a range in length of periods of five days to thirteen days, dis-
tributed as follows: five-day periods in each of five years, six-day
periods in each of four years, seven-day periods in each of two years,
eight-day periods in each of two years, and thirteen-day periods in each
of two years. The two thirteen-day periods are recorded for the years
1907, in which the average blooming period for all varieties was 9.93
days, and 1914, in which the average blooming period for all varieties
was 9.03 days. The mean daily temperature for the blooming period
in 1907 was 49.4°, in 1914 it was 60.4° ; a difference of 11°, altho the
blooming periods were the same. For the five years having each a five-
day period, the mean daily temperatures for the bloomirig periods
range from 47.4° in 1908 to 70.2° in 1913; an extreme difference of
22.8°.

The record for Domine shows equal flowering periods of six days
for the years 1906 and 1907, with a difference of 14.3° between the
mean daily temperatures of the periods for the two years. Similarly
the years 1903 and 1904 had each an eight-day period with a difference
between mean daily temperatures of 10°.

Variations in length of blooming period and relations to mean daily
temperatures such as have been cited are of common occurrence in all

1924] BLOOMING PERIODS OF APPLES . 143

varieties, but are more strikingly in evidence in some varieties than
in others.

The variations indicate clearly that there are more influences con-
trolling length of blooming period than can be ascribed to spring
weather conditions alone, and by their marked irregularity suggest
that the individual rather than the variety must be the basis from
which performance is considered. Trees of one variety growing to-
gether in a single row do not act alike; one starts earlier, pushes
stronger, and completes its period of bloom in less time than does its
neighbor. It is said of such a tree that it was in better condition than
the other, had more vigor, and thus was enabled to outclass its neigh-
bor. This is true, but 'when the question is asked as to why the tree
was in better condition and possessed of greater vigor, the answer is
not immediately forthcoming.

The two trees, Tolman and Whitney, had been neighbors for several
years, subjected to identical conditions so far as is now apparent. It
is necessary to go back several months or possibly years, investigate
relative activity in the various life processes, determine the perfor-
mance, and then possibly fail to find the existing cause or causes of the
differences recorded. The tree of greater vigor may be constitutionally
better than its neighbor, it may have had access to better food supply,
it may have more completely recovered from a fruiting effort, it may
not have been exhausted by fruit production the year previous while
its less vigorous neighbor may have been, or it may have escaped a
parasite that damaged its neighbor. Any of these or other possible
causes, singly or in combination, may have operated to bring about the
recorded difference, leaving no evidence of operation that is apparent
at the time bloom record is made.

144 BULLETIN No. 251 [May,

SUMMARY

Blooming marks what is perhaps the most critical period in the pro-
duction of a fruit crop, and success or failure of the crop is, in a large
measure, dependent upon spring weather conditions. Blooming rec-
ords of 106 varieties, each having a record of ten to sixteen years, are
considered for the sixteen-year period 1901-1916. Record for each
variety for each year includes dates of first bloom, full bloom, and
petals falling ; also, an estimate of the amount of bloom.

1. Flowering periods in Illinois are extremely variable from year
to year, but in general are shorter than flowering periods in Virginia,
New York, Oregon, and in England.

2. For the sixteen years the full flowering period for all varieties
has ranged from ten to twenty-two days, with an average of approxi-
mately sixteen days. The earliest date of an open flower was April
2 ; the latest date in the record of petals falling was May 21, placing
the extremes for the sixteen years fifty days apart.

3. Early Ripe, Zuzoff, and Oldenburg, while shifting relative posi-
tions from year to year, hold place as the earliest bloomers more con-
sistently than do other varieties. The latest blooming varieties weue
Repka Malenka and Wythe.

4. The blooming period in 1910 was a full month earlier than in
other years because of abnormally high temperatures during the month
of March.

5. Blooming periods for individual varieties have varied in length
between three and seventeen days, while average periods, for all years,
range from five to ten days. The average period for all varieties was
five days in 1901 and 1915 ; six days in each of two years, seven days
in each of six years, eight days in each of three years, nine days in
1903 and 1914, and ten days in 1907.

6. Extremes in length of blooming periods for individual varieties
may be near together in one year and widely separated in another
year. Thus in 1907 the shortest period was three days, the longest
seventeen days, and each number of days between three and seventeen
was represented by 1 to 13 varieties, giving fifteen time-groups for that
year, while in 1915 there was concentration into three groups of four-,
five-, and six-day periods.

7. Amount of bloom was estimated for each variety for each year
and from this record it appears that 68 varieties aggregate 151 records
of no bloom. For the full list of varieties there are 485 records of
bloom insufficient for a crop, and 1,060 records of a sufficient amount
or an excess of bloom. In other words, 621/2 percent of the records in-
dicate enough bloom and 37 1/2 percent represent deficiencies. Records
of very full or excess bloom aggregate only 92, or 5.42 percent of all

l'J24\ BLOOMING PERIODS OF APPLES 145

records, and these are distributed in eight of the sixteen years. Effort
to establish a relationship between amount of bloom and temperature
and moisture conditions prevailing during the period of bud forma-
tion in preceding summers developed such irregularities as to render
the existence of any definite relation doubtful.

8. In a general way mean temperatures during blooming periods
influence the length of periods and the distribution of varieties into
few or many time-groups, but there are enough irregularities to indi-
cate plainly that there are influences other than temperature that
operate to modify both length of period and range of distribution of
varieties.

9. Inconstancy in relative lengths of blooming periods is charac-
teristic of all varieties. Two varieties, one having a blooming period
twice as long as the other in a particular year, may exactly reverse
their relative positions in a succeeding year.

10. Influences affecting the blooming periods of apple trees are
extremely complex. Climatic conditions are important, but their in-
fluence is modified by physiological characteristics which are obscure,
difficult to isolate, and which operate to make it impossible to interpret
the vagaries of blooming phenomena from the bloom record alone.

UNIVERSITY OF ILLINOIS-URBANA