Experimental studies on the duration of life. XII. Influence of temperature during the larval period and adult life on the duration of the life of the imago of Drosophila melanogaster


Alpatov, W.W.; Pearl, R.

The American Naturalist 63(684): 37-67

1929


Summarizing, it has been shown in this paper that: 1. Drosophila reared at 18° C. is distinctly larger in a series of bodily dimensions than when reared at 28° C. Furthermore, the pigmentation is different in the two cases, notably with reference to the pigmentation of the seventh tergum. 2. Females were longer lived than males in all series of these experiments. The difference in duration of life between the two sexes diminishes as the temperature during imaginal life increases, in these experiments, regardless of the temperature during development. 3. As the temperature during imaginal life increases the duration of life decreases. The relationship between duration of life and temperature is exponential. 4. For the temperature range 18° to 28° used in these experiments the temperature coefficient for duration of life had, as average figures, the following values: Q10=2.07, μ=12717.3. 5. The relative or proportional influence of temperature upon body size and upon duration of life was, in these experiments, of the same order of magnitude. This fact furnishes confirmatory evidence to the theory that an important factor in determining the duration of life is the rate of energy expenditure during life.

EXPERIMENTAL
STUDIES
ON
THE
DURATION
OF
LIFE
XII.
INFLUENCE
OF
TEMPERATURE
DURING
THE
LARVAL
PERIOD
AND
ADULT
LIFE
ON
TIIE
DURA-
TION
OF
TIIE
LIFE
OF
TIIE
IMAGO
OF
DROSOPHILA
MELANOGASTER
1
W.
W.
ALPATOV2
AND
RAYMOND
PEARL
I
THE
origin
of
the
present
investigation
was
closely
connected
with
a
study
of
the
influence
of
temperature
on
physical
characteristics
of
Drosophila
conducted
by
one
of
the
authors
(W.
W.
A.)
in
the
Institute
for
Bio-
logical
Research
during
the
winter
of
1927-1928.
The
fl
ies
developed
under
different
temperatures
showed
such
pronounced
differences
in
the
body
size
as
well
as
in
the
coloration,
that
it
was
decided
to
test
their
longevity.
In
order
to
have
a
larger
material
for
discussion
it
was
thought
desirable
to
test
the
duration
of
life
of
"cold"
and
"warm"
fl
ies
in
three
incubators
running
at
different
temperatures.
As
a
result
of
this
expansion
of
the
problem
the
double
character
of
the
investigation
arises.
The
fi
rst
part
of
it
concerns
the
problem
of
the
influence
of
the
different
temperatures
on
the
duration
of
life
of
flies
genetically
and
phaenotypically
identical;
the
second
is
devoted
to
the
question
of
the
duration
of
imaginal
life
in
connection
with
different
conditions
of
larval
life,
i.e.,
development
under
different
temperatures.
Up
to
the
present
time
both
of
these
problems
have
been
only
occasionally
and
insufficiently
attacked
by
other
in-
vestigators,
and
have
not
been
worked
out
in
the
exact
form
adopted
by
Pearl
and
his
coworkers
from
the
ex-
perimental
as
well
as
the
statistical
points
of
view.
1
From
the
Institute
for
Biological
Research
of
the
Johns
Hopkins
Uni-
versity.
2
Research
Fellow
of
the
International
Education
Board.
37
38
THE
AMERICAN
NATURALIST
[VsL.
LXIII
Very
few
investigations
can
be
mentioned
in
connection
with
these
problems.
The
most
important
are
those
of
Loeb
(100),
Loeb
and
Northrop
(14)
and
Northrop
(17).
In
Loeb's
fi
rst
paper
on
the
influence
of
temperature
on
the
larvae
of
sea
urchins
the
temperature
coefficient
obtained
differs
considerably
from
the
usual
tempera-
ture
coefficients
for
different
biological
processes.
This
diversity
compelled
Loeb
to
emphasize
what
he
regarded
as
the
peculiar
nature
of
the
phenomenon
of
senescence,
in
comparison
with
other
biological
processes.
His
sec-
ond
paper
on
Drosophila
gave
quite
different
results
in
regard
to
the
magnitude
of
the
temperature
characteristic
of
the
duration
of
life,
showing
that
in
this
case
the
process
of
senescence
apparently
falls
within
the
limits
of
other
biological
processes
influenced
by
temperature.
At
the
present
time
we
know
that
the
temperature
coeffi-
cients
of
life
processes
vary
considerably
within
the
tem-
perature
scale.
Loeb's
experiments
on
the
mortality
of
larvae
of
the
sea
urchin
subjected
to
different
tempera-
tures
were
run
at
the
right
marginal
side
of
the
tem-
perature
scale,
and
the
data
can
not
be
compared
with
data
on
temperature
coefficients
of
other
biological
proc-
esses
obtained
in
the
middle
of
the
scale.
So
far
as
we
know,
Loeb
himself
never
tried
to
reconcile
his
findings
of
1908
with
his
later
observations
on
Drosophila.
Loeb
and
Northrop's
paper
(16)
contains
the
first
results
on
the
relation
of
the
duration
of
the
three
stages
of
life
of
Drosophila
at
different
temperatures.
These
authors
state
that
"as
far
as
our
present
experiments
go
the
ratio
of
the
duration
of
life
of
the
insect
to
the
dura-
tion
of
the
larval
stage
is
approximately
constant
for
all
temperatures
and
that
the
same
is
true
for
the
ratio
of
the
larval
to
the
pupal
stage."
Our
own
observations
upon
the
time
of
development
in
the
eggs
showed
that
there
is
an
extremely
wide
range
of
variation
depending
on
the
degree
of
development
in
which
the
egg
is
at
the
moment
of
oviposition.
At
a
temperature
28°
C.
the
range
of
No.
6E4]
THE
hITRATION
OF
LIFE
39
eclosion
is
between
a
few
hours
and
24
hours,
the
largest
number
of
eclosions
falling
in
the
interval
between
1S
and
24
hours
after
oviposition.
This
means
that
the
determination
of
the
beginning
of
the
larval
period,
even
if
the
period
of
oviposition
is
reduced
to
two
hours,
is
far
from
being
accurate
enough.
In
Loeb
and
Northrop's
experiments
(16)
the
procedure
is
even
less
precise.
The
authors
write
:
"Aseptic
flies
of
both
sexes
were
put
in
flasks
and
allowed
to
remain
15
hours
at
room
tempera-
ture,
during
which
time
a
number
of
eggs
were
laid.
. . .
The
larvae
hatch
in
a
few
hours
after
the
eggs
are
laid,
and
at
the
time
the
flies
were
removed
from
the
flask
most
of
the
larvae
had
already
hatched.
The
duration
of
life
of
the
larvae
was
reckoned
from
the
time
the
eggs
were
placed
in
the
incubator
to
the
time
the
pupae
were
formed."
Bonnier
(95)
used
a
much
more
accurate
method
for
the
determination
of
the
beginning
of
the
prepupal
period
of
the
development.
He
placed
fertil-
ized
flies
on
the
food
and
allowed
them
to
oviposit
only
two
hours.
The
result
was
that
Bonnier
succeeded
in
showing,
with
very
good
evidence,
that
"the
shortening
of
the
time
of
development
at
30°
C.
as
compared
with
the
time
at
25°
C.
is
much
more
pronounced
for
the
pupal
than
for
the
prepupal
stage."
It
can
be
seen
that
this
conclusion
is
just
opposite
to
Loeb
and
Northrop's
state-
ment
based,
as
we
believe,
on
a
technique
insufficiently
precise.
A
detailed
comparison
of
Loeb
and
Northrop's
data
with
our
own
will
be
made
later
on.
Northrop's
(17)
paper
has
also
a
certain
relation
to
our
problem.
He
attempted
to
prolong
the
larval
life
by
insufficient
feeding
and
to
fi
nd
if
such
a
prolongation
influences
the
duration
of
the
imaginal
life.
His
method
of
obtaining
larvae
was
even
less
precise
than
that
described
in
Loeb
and
Northrop's
paper.
This
consequently
produced
a
considerable
heterogeneity
in
respect
to
the
age
distribu-
tion
of
the
population
in
his
experimental
flasks.
The
question
of
the
correlation
between
the
length
of
larval
40
THE
AMERICAN
NATURALIST
1VoL.LXIII
and
imaginal
life
must
be
regarded
as
still
an
open
one.
We
can
not
refer
to
all
the
papers
devoted
to
the
in-
fluence
of
temperature
on
the
rate
of
development
of
other
insects
than
Drosophila.
They
are
very
numerous
but
only
a
few
can
be
mentioned
as
fulfilling
the
require-
ments
of
an
exact
experimental
work.
To
this
category
belong
the
papers
of
Peairs
(102),
Titschack
(105),
Krogh
(99),
Janisch
(97),
and
Bliss
(94).
In
connection
with
our
second
problem
—the
size
of
the
body
and
the
duration
of
life
—only
one
paper
may
be
mentioned.
This
is
the
article
of
Titschack
(106),
cited
from
Janisch
(98).
This
author
found
that
the
larger
moths
had
also
a
longer
duration
of
life.
II
The
flies
used
in
our
experiments
belong
to
the
wild
culture
of
Drosophila
melanogaster
known
as
Line
107
(Pearl
and
Parker,
32).
On
December
1,
1927,
the
par-
ents
of
experimental
fl
ies
were
taken
from
sixteen
stock
bottles,
each
starting
with
fi
ve
pairs
of
fl
ies.
One
hun-
dred
bottles
were
populated
with
fi
ve
pairs
each
and
divided
into
two
parts
(50
and
50)
;
one
series
was
placed
in
an
incubator
running
at
18°
C.,
the
other
in
an
incuba-
tor
at
28°
C.
The
parents
were
kept
on
the
food
for
six
days.
From
the
fi
fty
bottles
kept
at
28°
C.
in
47
bottles
fl
ies
had
emerged
on
December
9.
On
December
10
the
28°
C.
bottles
gave
material
representing
the
"warm"
flies.
The
flies
were
collected
on
December
10.
On
the
day
before
all
the
fl
ies
which
had
then
emerged
were
shaken
from
the
28°
C.
bottles.
Therefore
the
age
of
the
flies
at
the
moment
of
beginning
the
duration
of
life
experiment
was
between
0
and
24
hours.
The
fl
ies
were
placed
in
one
-ounce
bottles,
25
males
and
25
females
in
each.
The
bottles
were
divided
into
three
groups
and
placed
in
incubators
running
respectively
at
18°,
25°,
and
28°
C.
The
emergence
of
fl
ies
in
the
bottles
kept
at
18°
C.
was
naturally
more
delayed.
Emerged
fl
ies
were
observed
No.
684]
THE
DURATION
OF
LIFE
41
in
15
bottles
on
December
19,
and
in
38
on
December
20.
The
"
cold"
fl
ies
which
emerged
on
December
21
furnished
the
material
for
incubators
running
at
18°
and
25°
C.
Those
which
emerged
on
December
22
were
put
in
the
incubator
running
at
28°
C.
In
calculating
the
average
duration
of
life
we
did
not
count
this
0
-24
-hour
period,
calculating
the
age
of
our
fl
ies
from
the
moment
of
put-
ting
the
fl
ies
in
the
ounce
bottles.
In
this
work
we
used
for
the
temperatures
of
28°
and
25°
ordinary
electric
incubators,
and
for
the
temperature
of
18°
C.
a
Hearson
low
temperature
incubator,
with
an
ice
-box
and
electrical
heating.
We
were
obliged
to
use
25°
C.
as
an
intermediate
temperature
because
of
the
difficulty
in
running
an
electric
incubator
at
23°
C.,
this
temperature
being
too
close
to
the
room
temperature
of
the
laboratory.
At
that
time
we
did
not
have
a
second
Hearson
incubator,
which
would
have
allowed
us
to
have
the
desired
temperature
of
23°
C.
Besides
having
accu-
rate
chemical
thermometers
the
incubators
were
provided
with
thermographs.
The
temperature
records
were
taken
twice
a
day,
at
9
A.
M.
and
5
P.
M.
We
can
not
say
that
the
temperature
in
the
above
mentioned
incubators
can
be
kept
extremely
constant.
Calculating
biometrical
con-
stants
for
morning
temperatures
of
the
fi
rst
100
days
of
the
experiment
we
get
the
data
shown
in
Table
1.
TABLE
1
CONSTANTS
FOIL
TEMPERATURE
VARIATION
IN
THE
THREE
INCUBATORS
Constant
18°
25°
28°
Mean
,,.
,,,,,
„,,,..„,„
,,,,,
,.„
,,,,,
.,..:
....
.,
.
n....,
18.04°
±
.04
25.13°
±
.03
28.00°
±
.04
Standard
deviation.
0.65°
0.52° 0.62°
Coefficient
of
variation_
3.58
±
.17
2.05
±
.10
2.22
±
.11
Limits
16.1°
20.0°
23.1°
27.3°
25.8°-30.3°
It
must
be
noted
that
in
the
incubator
running
at
18°
C.
the
temperature
fell
three
times
to
8°-10°
C.,
due
to
faulty
operation
of
the
regulatory
apparatus.
According
to
the
42
THE
AMERICAN
NATURALIST
[VOL.
LXIII
thermographic
records
this
undercooling
took
place
only
for
a
few
hours
and
apparently
did
not
influence
the
vitality
of
the
experimental
animals.
However,
these
three
exceptional
cases
are
not
included
in
the
data
which
gave
us
the
material
for
the
calculation
of
Table
1.
It
can
be
seen
that
although
the
variation
of
temperature
is
rather
broad,
the
extreme
deviations
occurred
so
rarely
that
the
averages
are
very
close
to
the
desired
tempera-
tures.
The
greater
part
of
the
data
is
concentrated,
In
fact,
within
±
limits
from
the
average
temperature.
This
fact
is
made
apparent
biometrically
by
the
very
low
coefficients
of
variation.
III
Turning
attention
first
to
the
influence
of
temperature
on
the
time
of
development
of
Drosophila,
it
is
to
be
noted
that
this
phenomenon
exhibits
considerable
time
varia-
tion.
We
observed
two
points
:
fi
rst,
the
moment
of
putting
the
parental
flies
on
the
food;
and,
second,
the
average
time
of
appearance
of
the
emerged
flies
in
the
bottles.
The
same
data
were
recorded
concerning
the
development
of
flies
which
were
bred
for
the
purpose
of
a
biometrical
study.
As
can
be
seen
from
Table
2,
the
TABLE
2
AVERAGE
TIME
(IN
DAYS)
FROM
TIIE
BEGINNING
OF
THE
CULTURE
UNTIL
THE
EMERGENCE
OF
THE
FLIES
Material
18°
2
28°
3
Our
material
prepared
for
duration
of
life
experiment
19.00
±
.08
8.06
±
.02
Our
material
for
mea-
surements
.
16.25
±
.15
9.58
±
.04
8.55
±
.12
Bonnier
's
data
-
9.6
7.6
total
duration
of
development
at
28°
C.
is
approximately
half
as
long
as
that
at
18°
C.
No.
684]
THE
DURATION
OF
LIFE
43
The
physical
characteristics
of
flies
developed
at
dif-
ferent
temperatures
have
been
measured
on
individuals
originating
from
a
specially
organized
experiment.
The
history
of
the
animals
of
that
experiment
is
as
follows.
On
October
20,
1927,
20
pairs
(g'
and
were
taken
from
the
mass
culture
of
the
Wild
Line
107,
and
put
into
20
half-pint
bottles
with
100
cc
synthetic
food.
The
most
fertile
bottles
(according
to
the
fi
rst
and
second
days
of
emergence)
gave
origin
for
brother
and
sister
co
atings
(g
and
)
in
which
virgin
females
were
used.
From
one
of
the
most
fertile
of
these
brother
y,
sister
bottles,
flies
were
taken
and
placed
(5e
and
5
?)
in
three
FIG.
1.
Points
defining
the
measurements
of
the
length
of
the
fi
fth
and
sixth
tergum.
bottles
each
and
placed
in
the
incubator
with
correspond-
ing
temperature.
The
fi
rst
generation
of
flies
developed
under
different
conditions
was
measured,
50
males
and
50
females
being
taken
for
each
temperature.
The
fol-
lowing
measurements
were
chosen:
(i)
Width
of
the
head
(including
the
compound
eyes).
(2)
Length
of
the
fi
fth
tergum
(assuming
that
the
first
tergum
is
mor-
phologically
a
double
segment)
(Fig.
1).
(3)
Length
of
the
sixth
,
tergum
(Fig.
1).
(4)
Length
of
the
femur
(middle
pair
of
legs)
(Fig.
2).
(3)
Length
of
the
tibia
(middle
pair
of
legs)
(Fig.
2).
(6)
Distance
between
the
anterior
and
posterior
crossveins
(Fig.
3).
(7)
Length
of
the
posterior
crossvein
(Fig.
3).
(8)
Length
of
the
wing
(Fig.
3).
(9)
Width
of
the
wing
(Fig.
3).
The
measurements
were
made
with
the
aid
of
a
Spencer
screw
ocular
micrometer;
the
fi
rst
seven
measurements
44
THE
AMERICAN
NATURALIST
[VoL.
LXIII
were
made
with
a
Leitz
No.
3
objective,
and
the
eighth
and
ninth
with
a
Reichert
No.
1
objective,
in
both
cases
on
a
Spencer
microscope.
The
material
was
preserved
in
70
per
cent.
alcohol
and
before
measuring
it
was
mounted
under
cover
glasses
in
glycerin.
C
rt
FIG.
2.
Points
defining
the
measurements
of
the
length
of
the
tibia
(A
—B)
and
the
femur
(A
—C)
of
the
middle
pair
of
legs.
The
accompanying
Tables
3
and
4
show
the
differences
which
exist
between
flies
belonging
to
the
different
experi-
mental
series.
In
the
central
column
of
these
tables
R
denotes
the
ratio
of
the
difference
between
correspond-
ing
means
to
its
probable
error.
The
number
of
char
-
Pm.
3.
Measurements
of
the
wing.
AB,
length
of
the
wing.
EF,
distance
between
the
anterior
and
posterior
crossveins.
CD,
width
of
the
wing.
EG,
length
of
the
posterior
crossvein.
acters
measured
on
males
was
reduced.
Characters
2
and
3
were
omitted
because
of
the
difficulties
in
straight-
ening
the
tergums
on
the
slide.
The
coefficients
of
variation
in
these
two
characters
in
the
females
are
extra-
TABLE
3
BIOMETRICAL
CONSTANTS
FOR
FEMALES
OP
DROSOPIIILA
REARED
AT
18°
AND
28°
C.
Characters
development
a
9
t
at
18°
C.
Differences
9
development
at
28°
Standard
deviation
C.
Mean
Standard
deviation
C.
of
V.
in
means
Mean
C.
of
V.
1
0.9054
±..0021*
0.0216
2.39
±
.16
0.0627
±
.0030
0.8427
±
.0022
0.0228
2.71
±
.18
R=20.9
2
0.2834
±
.0021
0.0219
7.73
±
.52
0.0003
±
.0027
0.2831
±..0017
0.0173
6.11
±
.41
R=
0.1
3
0.2913
±
.0020
0.0207
7.11
±
.48
0.0064
±
.0028
0.2846
±..0020
0.0204
7.52
±
.51
R=
2.4
4
0.7070
±
.0014
0.0146
2.07
±
.14
0.0402
±
.0021
0.6668
±
.0015
0.0156
2.34
±..16
R=
19.1
5
0.7246
±
.0016
0.0172
2.37
±
.16
0.0433
±
.0021
0.6813
±
.0013
0.0135
1.98
±
.13
R=20.6
6
0.4954
±..0017
0.0180
3.63
±..24
0.0614
±
.0024
0.4340
±
.0016
0.0167
3.84
±
.26
R
=
25.6
7
0.1990
±
.0008
0.0083
4.17
±..28
0.0178
±..0012
0.1812
±..0009
0.0092
5.08
±
.34
R=14.8
8
1.8832
±
.0033
0.0343
1.82
±
.12
0.1837
±..0045
1.6995
±
.0030
0.0312
1.84
±
.12
R=40.8
9
1.0690
±..0017
0.0187
1.75
±
.12
0.1007
±
.0024
0.9683
±
.
.0017
0.0169
1.75
±
.12
R
=
42.0
*
All
measurements
are
in
millimeter
s.
Li0
Rouyn
JILL
TABLE
4
BIOMETRICAL
CONSTANTS
FOR
MALES
OF
DROSOPHILA.
REARED
AT
18°
AND
28°
C.
Characters
S
development
at
18°
C.
C.
of
V.
Differences
in
means
development
at
28°
C.
Mean
Standard.
deviation
Mean
Standard
deviation
C.
of
V.
1
0.8290
±
.0021°
0.0218
2.63
.18
0.0248
±
.0030
0.8042
±
.0021
0.0220
2.74
-
±.
.18
R
=
8.3
4
0.6766
±
.0010
0.0105
1.55
.10
0.0472
-
±.
.0017
0.6294
±
.0014
0.0151
2.40
-
±.
.16
=
27.8
5
0.6937
±
.0013
0.0135
1.95
-
±.13
0.0477
±
.0021
0.6460
±
.0016
0.0165
2.55
±
.17
=
22.7
8
1.7310
±
.0018
0.0190
1.10
±
.07
0.2324
±
.0038
1.4986
±
.0034
0.0352
2.35
-
±.
.16
=
61.2
9
1.0194
±
.0014
0.0151
1.49
±
.10
0.1324
±
.0024
0.8800
±
.0019
0.0201
2.28
±
.13
=
55.2
:F
All
measurements
are
in
millimeters.
No.
684]
THE
DURATION
OF
LIFE
47
ordinarily
high.
This
is
due
without
doubt,
not
primarily
to
the
organic
variation
of
that
particular
characteristic,
but
to
the
distortions
arising
during
the
dissection
and
mounting
of
the
object
on
the
slide.
It
is
evident
from
Tables
3
and
4
that
the
flies,
females
as
well
as
males,
developed
at
the
low
temperature
are
significantly
larger
in
all
characters
other
than
2
and
3
than
are
the
warm
temperature
ones.
Besides
these
differences
in the
dimensions
of
the
body
another
very
interesting
peculiarity
of
the
cold
temperature
fl
ies
must
be
mentioned.
This
is
in
the
development
of
the
black
spots
on
the
tergum
and
at
the
base
of
the
wings
(see
Fig.
4).
It
is
to
be
noted
that
the
seventh
tergal
spot,
4
/MM.
FIG.
4.
Side
view
on
the
abdomen
of
two
female
specimens
of
Drosophila,
melanogaster.
Above
is
a
specimen
from
the
28°
culture,
below
one
from
the
18°
culture.
The
drawings
were
made
with
a
camera
lucida
to
the
same
degree
of
magnification.
TABLE
5
SURVIVORSHIP
DISTRIBUTIONS
OF
FLIES
UNDER
DIFFERENT
CONDITIONS
AS
TO
THE
TEMPERATURE
OF
DEVELOPMENT
AND
IMAGINAL
LIFE
Temperature
of
development
Temperature
of
imaginal
life
18°
18°
18°
1
2
0
0
28°
1
°
28°
28°
25°
25°
28°
Days
a
9
a
9
0—
4
1,000*
1,000*
1,000* 1,000*
1,000*
1,000* 1,000* 1,000*
1,000*
1,000* 1,000* 1,000*
5—
9
.....
996
986
986
986
992
966 976
970
978
978
982
968
10—
14
992
982
982
984
986
964
974
966
966 956 966
962
15—
19.......,
980
974
954
958
878
928
962
958
880
909
884
890
20—
24
942
968
774
872
522
710
918
940
502
833
488
768
25—
29..__....:
906
902
622
794
316
620
862
908
326
712
332
646
30—
34,
834
862
468
704
152
546
584
850
176
583
188
456
35—
39
...
790
842
116
620
30
416
456
820
78
478
108
252
40—
44.
676
822
30
544
2
304
310
796
46
379
48
150
45—
49
,.
484
806
14
448
2
224
236
750
24
302
16
78
50—
54........
.-,
256
732
10
310
112
136
682
14
224
4
26
55—
59..
178
674
4
220
32
56
634
6
175
2
8
60—
64
80
602
2
144
8
12
550
2
72
65—
69
40
554
m
,.
78
4
8
488
36
70—
74
..,„,,
20
478
24
4
360
12
75—
79.
12
400
6
t.--,
,
304
80—
84...
. .
4
372
260
85—
89
,
..
2
314
224
90—
94
....
286
210
95—
99
.....
.
..
230
w.
188
100-104
,„„,.,..,.,
180
170
105-109._....._
154
M
142
110-114...,„„_.„
134
96
115-119..
104
82
THE
AMERICAN
NATURAL
0
TABLE
5
-(Continued)
Temperature
of
development
Temperature
of
imaginal
life
18°
18°
18°
25°
18°
28°
28°
1S°
28°
25°
28°
2S°
Days
8
a
4
8
a
9
a
9
a
9
120-124
.
7S
60
125-129
68
50
130-134
.
52
38
135-139
30
26
140-144
18
12
145-149.
S
8
150-154
6 6
155-159
2
b..
2
160-164
,
2
Mean
43.46
±.40
70.61
±.98
26.81
±.24
40.96
±.50
21.90
±.21
30.67
±.43
34.97
±.37
65.25
±.96
22.49
±.27
35.75
±.50
22.59
±.27
28.52
±.34
Standard
de-
viation
C.
of
IT.
Absolute
num-
ber
of
flies
13.08
32.53
8.129
16.57
6.909
14.03
12.33
31.809
8.981
16.34
5.862
11.12
30.10
493
46.07
±1.17
30.32
±.70
40.46
±1.00
31.54
±.74
45.76
±1.17
35.25
±.84
48.75
±1.27
39.93
+.98
44.46
±1.13
39.23
±.97
38.98
±.96
499
501
497
497
493
496 496
499
492
491
493
*
The
fi
gures
in
these
columns
give
the
number
of
survivors
on
the
fi
rst
day
of
the
age
interval
indicated
in
the
left-hand
column
of
the
table,
headed
"Days."
CA
41
,
JO
NOLEVffila
gIII
50
THE
AMERICAN
NATURALIST
LICHT
which
is
completely
invisible
in
warm
(28°)
temperature
females
appears
in
a
perfectly
developed
form
only
in
flies
reared
at
18°
0.
Twenty-five
degrees
flies
have
only
traces
of
black
pigment
in
the
seventh
tergum.
IV
Table
5
presents
the
basic
data
of
the
experiments
on
duration
of
life.
The
numerators
of
the
fractions
head-
ing
the
columns
indicate
the
temperature
during
de-
velopment,.
the
denominators
that
during
imaginal
life.
The
number
of
survivors
is
calculated
on
the
basis
of
1,000
individuals
at
the
moment
of
putting
the
flies
in
the
one
-ounce
bottles.
The
first
point
to
which
attention
may
be
called
is
the
difference
between
the
sexes
in
mean
duration
of
life.
In
all
six
series
the
females
have
a
significantly
longer
duration
of
life
than
the
males.
The
differences,
with
their
probable
errors,
are
shown
in
Table
6.
DIFFERENCES
Series
TABLE
BETWEEN
THE
SEXES
IN
MEAN
DURATION
OF
LIFE
Female
mean
minus
male
mean
(days)
Ratio
Diff./P.E.
Diff.
18718°
+27.15
±
L06
25.6
18725
0
+
14.15
±
.55
25.7
18728°
+
8.77
±
.48
18.3
28718
0
+
30.28
±
1.08
29.4
28725°
+
13.26
±
.57
23.3
28
0
/28
0
+
5.93
±
.43
13.8
From
Table
6
it
is
obvious
that
all
the
sex
differences
are
significant,
by
large
margins.
In
the
case
where
the
two
sexes
are
nearest
together
in
mean
duration
of
life
(28°/28°)
the
difference
between
the
two
means
is
nearly
14
times
its
probable
error.
This
result
confirms
what
has
been
found
in
the
earlier
life
table
work
on
Dro-
sophila.
In
the
definitive
life
tables
calculated
from
critically
controlled
ad
hoc
material
for
this
form
(Pearl
No.
684]
THE
DURATION
OF
LIFE
51
and
Parker,
66),
a
sex
difference
in
the
same
direction
was
observed
in
both
wild
type
and
vestigial
fl
ies,
the
difference
being
more
marked
in
the
case
of
the
yes-
tigials.
Regarding
this
matter
Pearl
(103)
said:
"It
seems
to
be
a
rather
general
phenomenon,
among
groups
of
or-
ganisms
in
-
which
there
is
sexual
dimorphism,
for
males
to
be
shorter
-lived
on
the
average
than
females,
but
sta-
tistically
adequate
quantitative
information
about
dura-
tion
of
life
is
so
meager
that
any
such
generalization
would
be
premature
at
the
present
time.
Such
cases
as
those
of
Dinophilus
apatris
and
various
rotifers
dis-
cussed
by
Korschelt,
in
which
the
male
is
dwarfed
and
obviously
deficiently
organized,
and
at
the
same
time
is
short-lived,
as
compared
with
the
female,
are
perhaps
to
be
regarded
as
extreme
illustrations
of
the
dependence
of
duration
of
life
upon
bodily
organization
and
pat-
tern,
a
point
which
will
be
more
fully
discussed
in
later
chapters.
But
these
cases
are
not
entirely
probative
evidence
in
support
of
a
hypothesis
that
the
organiza-
tion
or
pattern
differences
implicate
in
normal
sexual
dimorphism
of
the
kind
and
degree
seen
in
man,
for
ex-
ample,
have
generally
as
one
of
their
normal
expressions
a
shorter
average
duration
of
life
in
the
male.
Kor-
schelt
apparently
inclines
to
the
opinion
that
such
a
rela-
tionship
is
general,
but
makes
the
same
point
as
is
em-
phasized
above,
that
the
data
available
are
insufficient
to
settle
the
question.
It
is
of
some'
interest
to
note
that
Blunck
fi
nds
the
average
duration
of
life
of
female
beetles
(Dytiscus)
greater
than
that
of
males,
and
also
that
Labitte's
observations
on
the
duration
of
life
in
beetles
generally
show
the
same
thing."
The
present
experiments
show
that,
in
the
case
of
wild
type
Drosophila
the
sense
of
this
relationship
between
the
sexes
is
not
disturbed
by
the
temperature
at
which
they
are
reared
or
in
which
the
imagoes
live.
But
from
Table
6
it
also
plainly
appears that
the
dif-
ference
between
the
two
sexes
in
mean
duration
of
life
52
THE
AMERICAN
NATURALIST
[Vol,.
LXIII
diminishes
as
the
temperature
during
imaginal
life
in-
creases,
regardless
of
the
temperature
during
develop-
ment.
This
result
is
shown
graphically
in
Fig.
5.
Reared
at
_28
°
.,
Ire
%.•
\
\
18°
`28°
t
i
t
f
I
I
Ag
e
20
°
22
°
24
0
26
°
28
°
knaginal
Life
Temperature
FIG.
5.
Showing
the
decrease
in
the
excess
of
female
over
male
mean
duration
of
life,
with
increasing
temperature
during
imaginal
life.
V
There
are
obviously
two
distinct
questions
which
want
discussion
relative
to
the
influence
of
temperature
on
duration
of
life.
They
may
be
formulated
in
the
fol-
lowing
manner
:
1.
What
is
the
effect
of
the
temperature
during
imag-
inal
life
upon
the
duration
of
life
of
Drosophila?
No.
684]
THE
DURATION
OF
LIFE
53
2.
What
is
the
effect
of
the
temperature
during
de-
velopment
upon
the
subsequent
duration
of
life
as
imago
To
answer
the
first
question
requires
that
the
data
be
exhibited
in
such
way
as
to
show
the
change
in
mean
duration
of
life
associated
with
the
same
temperature
during
development
but
different
temperatures
during
imaginal
life.
This
is
done
in
Figs.
6
and
7.
50
4(0
30
20
`•-•
(1
Z
/
0
fr
-e/
040Q,
e2
/
4
/
8
0
Males
/8°
25°
23°
Temperature
During
/mag/nal
Life
FIG.
6.
Graph
showing
the
average
duration
of
life
of
males
of
Drosophila
kept
at
different
temperatures.
The
circles
refer
to
flies
developed
at
18°,
the
triangles
to
those
developed
at
28°.
In
Figs.
6
and
7
are
shown
the
results
of
fi
tting
to
the
observed
mean
durations
of
life
given
in
Table
5
an
ex-
ponential
equation
of
the
type
54
THE
AMERICAN
NATURALIST
[Vol,.
LXIII
70
/0
Females
/3
°
25'
2
Temperature
During
/mg/nal
Life
FIG.
7.
The
same
as
Fig.
6,
but
for
females.
=
a
in
which
y
denotes
mean
duration
of
life,
and
t
tempera-
ture
during
imaginal
life,
in
degrees
Centigrade.
The
equations
for
the
males
and
females
developed
at
18°
C.,
and
that
for
females
at
28°
C.,
have
been
calcu-
No.
684]
THE
DURATION
OF
LIFE
55
lated
on
the
basis
of
the
observed
durations
of
life
at
18°
and
28'.
On
account
of
the
abnormally
high
duration
of
life
of
males
in
the
series
28°/28°
the
equation
for
this
series
was
computed
from
the
observations
at
18'
and
25°.
From
Figs.
6
and
7
it
is
apparent
that
the
present
ex-
periments
confirm
the
principal
result
of
Loeb
and
Northrop's
studies,
namely
that
as
the
temperature
dur-
ing
imaginal
life
is
higher
Drosophila
lives,
on
the
aver-
age,
a
shorter
time.
This
is
true
of
both
sexes,
and
for
both
series
of
developmental
temperatures.
The
problem
of
a
rational
explanation
of
this
tempera-
ture
effect
upon
duration
of
life
is
an
interesting
one.
The
type
of
explanation
which
has
been
most
in
favor
since
Loeb
discussed
the
matter
may
be
characterized
as
basically
a
chemical
one.
Loeb
and
Northrop
concluded
that:
"The
observations
on
the
temperature
coefficient
for
the
duration
of
life
suggest
that
this
duration
is
de-
termined
by
the
production
of
a
substance
leading
to
old
age
and
natural
death."
The
reasoning
back
of
this
conclusion
is
that
temperature
controls
the
rate
of
pro-
duction
of
these
hypothetical
chemical
substances.
An
alternative
explanation
is
biological
in
its
essence,
rather
than
chemical.
It
is
a
simple
fact
of
observation
that
the
total
activity
(movement,
etc.)
of
Drosophila
is
greater
in
higher
temperatures
and
less
in
lower
tem-
peratures.
Now
Pearl,
in
a
considerable
amount
of
recent
work
(ef.
103,
and
other
publications
there
cited,
and
104),
and
MacArthur
and
Baillie
(101),
have
shown
that
total
duration
of
life
varies
inversely
as
the
rate
of
energy
expenditure
in
living
(growth,
muscular
move-
ment,
etc.).
Wheeler
(107)
is
of
the
opinion
that
in
the
inverse
relation
between
rate
of
living
and
duration
of
life
is
probably
to
be
found
the
physiological
basis
for
the
lengthening
of
the
adult
life
in
social
insects.
He
points
out
that
"all
the
subsocial
and
social
insects
live
in
small
cavities
of
the
soil
or
wood,
in
hives
or,
in
the
more
exceptional
cases
of
social
wasps
and
certain
trop-
56
THE
AMERICAN
NATURALIST
[Vol,.
LXIII
ical
ants,
in
the
cavities
of
carton
nests.
The
environ-
ment
is,
therefore,
one
which
restricts
or
inhibits
muscu-
lar
movement
and
is
dark,
poor
in
oxygen,
and
of
rather
low
and
uniform
temperature.
All
of
these
conditions
would
necessarily
favor
a
lowered
rate
of
metabolism
and
activity
and
an
accumulation
of
fat
in
the
insect
body.
The
queens,
or
mothers
of
insect
societies
cer-
tainly
impress
one
as
having
acquired
their
physiolog-
ical
and
some
of
their
morphological
peculiarities
as
re-
sponses
to
just
such
an
environment,
for
they
are
very
sluggish
and
tend
to
lose
the
powers
of
flight
(Meli-
poninae)
or
even
the
wings
(ants
and
termites)
and
to
acquire
an
accentuated
anabolism
as
shown
in
the
accu-
mulation
of
fat
and
of
yolk
-laden
eggs."
Furthermore,
Wheeler
notes
that:
"Certainly
the
life
-span
of
the
three
castes
of
ants
and
social
bees
would
seem
to
be
roughly
proportional
to
their
respective
expenditures
of
en-
ergy."
It
is
in
accord
with
the
results
of
the
body
of
carefully
controlled
experimental
work
cited
i1
(103)
and
(104)
to
suppose
that
the
reason
why
Drosophila
has
a
shorter
duration
of
imaginal
life
at
higher
temperatures
is
pri-
marily
simply
because
it
is
more
active
at
those
tempera-
tures;
or,
in
other
words,
has
a
higher
"rate
of
living,"
and
consequently
a
shorter
absolute
duration
of
life.
To
us
such
an
explanation
seems
inherently
more
prob-
able
than
the
hypothetical
chemical
substances
postulated
by
Loeb
and
Northrop,
but
we
have
no
desire
to
press
the
point
now,
hoping
with
the
passage
of
time
to
collect
further
experimental
evidence
which
will
make
possible
a
critical
quantitative
discrimination
between
the
two
alternatives.
Figs.
6
and
7
show
an
interesting
difference
between
the
sexes
in
respect
of
the
influence
of
temperature
upon
duration
of
life.
The
duration
of
life
of
the
females
is
proportionately
more
shortened
by
high
temperatures
during
imaginal
life
than
is
that
of
the
males.
Thus,
taking
flies
developed
at
18°,
we
see
that
whereas
the
No.
684]
THE
DURATION
OF
LIFE
57
duration
of
life
of
the
males
living
as
imagoes
at
28°
is
shortened
below
that
of
males
living
at
18°
by
49.6
per
cent.
of
the
mean
duration
of
life
of
the
latter,
the
corre-
sponding
fi
gure
for
the
females
is
56.6
per
cent.
VI
Plotting
the
duration
of
life
at
different
temperatures
on
arithlog
paper
we
fi
nd
that,
with
the
exception
of
males
in
the
28°/28°
series,
the
data
give
a
straight
linear
distribution.
We
are
therefore
justified
in
concluding
that
within
the
temperature
limits
here
used,
the
dura-
tion
of
life
is
an
exponential
function
of
the
temperature.
This
result
permits
the
calculation
of
the
temperature
coefficients
commonly
used
in
biochemical
and
biological
literature,
namely
van't
Hoff's
Q
10
and
Arrhenius's
p.
This
was
done
(see
Table
7)
for
our
own
material
and
for
the
material
of
Loeb
and
Northrop
(see
Table
8).
Inspection
of
the
values
of
Q
and
p
collected
in
Table
7
shows
that
there
is
a
pronounced
difference
between
the
sexes
in
respect
to
these
constants.
The
females
have
higher
temperature
increments
than
the
males.
Graphically
it
is
seen
in
a
sharper
decline
of
the
female
curves
of
Fig.
7,
as
compared
with
the
males
shown
in
Fig.
6.
In
general,
the
temperature
coefficients
obtained
in
this
work
are
lower
than
those
found
by
Loeb
and
Northrop
for
Drosophila.
What
the
significance,
if
any,
of
this
fact
may
be,
we
are
not
able
to
say.
•ur
mean
value
for
Q„
of
2.07
is
at
the
classical
spot
for
a
purely
chemical
reaction
in
a
homogeneous
system,
according
to
the
RGT-rule
as
enunciated
by
van't
Hoff.
But
it
would
be
unwarranted,
we
think,
to
conclude
from
this
fact
either
that
duration
of
life
is
determined
by
a
purely
chemical
reaction,
or
that
Drosophila
is
a
chemically
homogeneous
system.
In
fact
we
are
tolerably
certain
that
neither
of
these
conclusions
is,
in
fact,
true,
in
spite
of
the
value
Q
10
=
2.07.
We
are
in
agreement
with
the
position
taken
by
Hober
(96)
to
the
effect
that
great
cau-
58
THE
AMERICAN
NATURALIST
[Vol,
LXIII
TABLE
7
TEMPERATURE
COEFFICIENTS
Q
AND
It
FOR
THE
DURATION
OF
LIFE
OF
Drosophila
velanogaster
.
10
Flies
developed
at
18°
Sea
18
0
-28
0
--
18°-25°
-__
25°-28°
8
--,-
1.96
1.98
1.99
9
2.30
2.18
2.62
Flies
developed
at
28°
1.77
1.88
..
:
9
2.29
2.36
2.12
log
Kt.,
-
log
K
t
.
4.605
u
-
T
-1'
1'
'I
'
2
1
Flies
developed
at
18°
12005.7
11968.3
14607.7
13492.2
Flies
developed
at
28°
12096.1
17299.8
7695.2
10936.5
14497.9
1007.5
13510.7
TABLE
S
TEMPERATURE
COEFFICIENTS
Q
AN)
[1,
FOR
THE
DURATION
OF
LIFE
OF
Drosophila
melanogaster
Series
Q
ILL
18
0
-28°
(our
material
sununed
up)
2.07
12717.3
▪1
0°-20°
2.99
-
17894.1
Loeh's
data
...
.
15°-25°
3.24
20866.4
20
0
-30
0
2.96
20556.5
Lion
must
be
used
in
drawing
biological
conclusions
from
the
numerical
values
of
temperature
coefficients.
No.
684]
THE
DURATION
OF
LIFE
59
VII
We
may
turn
now
to
the
second
problem
stated
above,
namely
the
effect
of
the
temperature
during
embryonic,
larval
and
pupal
development
upon
the
duration
of
imaginal
life.
The
data
for
mean
duration
of
life
from
Table
5
supra
are
so
arranged
in
Table
9
as
to
give
the
answer.
The
results
are
shown
graphically
in
Figs.
S
to
1:3.
/000
One
at
ir
/00
/0
/Vales
imaginal
Tempera/are
/8°
0
/5
JO
4
,
5
60
75
90
Age
in
days
FIG.
S.
Survivorship
fl
ues
of
males
kept
during
imaginal
life
at
18°
C.
Ill
this
figure,
as
well
as
ill
Figs.
9,
10, 11,
12
and
13,
the
coutiuuous
Mlle
represents
fl
ies
developed
at
18°
C.,
the
brokeu
line
those
developed
at
28°
C.
/os
From
Table
9
and
the
diagrams
it
is,
in
the
fi
rst
place,
apparent
that
the
temperature
at
which
development
(embryonic,
larval
and
pupal)
takes
place
affects
the
60
THE
AMERICAN
NATURALIST
[VoL.
LXIII
/000
/0
&gat
28°',
Dem
at
18°
Ma/es
/magma/
Tempera/are
25°
0
/..7
30
45
00
75
Age
in
days
FIG.
9.
Survivorship
lines
of
males
kept
during
imaginal
life
at
25°
C.
TABLE
9
THE
EFFECT
OF
TEMPERATURE
DURING
DEVELOPMENT
UPON
THE
DURATION
OF
IMAGINAL
LIFE
AT
THREE
CONSTANT
TEMPERATURES
Temper-
ature
during
imaginal
life
Males
Mean Mean
duration
duration
of
life
of
life
of
fl
ies
of
fl
ies
reared
reared
at
18°
at
28°
Difference
Diff.
P.E.
Duff.
18°
43.46
±
.40
34.97
.37
+
8.49
±
.54
15.7
25°
26.81
±
.24
22.49
±
.27
+
4.32
±
.36
12.0
28°
21.90
±
.21
22.59
±
.27
-
.69
±
.34
2.02
Females
18°
70.61
±-
.98
65.25
.96
+
5.36
±-
1.37
3.91
25°
40.96
±
.50
35.75
.50
+
5.21
±
.71
7.3
28°
30.67
.43
28.52
±
.34
+2.15
±
.55
3.9
No.
684]
THE
DURATION
OF
LIFE
61
/000
-
/00
Deglat/8°
Deuat.28°
Males
baginallitnpetatare
1-
11
O
/3
JO
//3
60
75
Rge
in
days
FIG.
10.
Survivorship
lines
of
males
kept
during
imaginal
life
at
2S°
C.
duration
of
subsequent
imaginal
life,
regardless
of
the
temperature
at
which
the
latter
lived.
A
high
tempera-
ture
(28°)
during
development
shortens
the
duration
of
the
subsequent
imaginal
life
at
all
temperatures
and
in
both
sexes.
The
only
exception
to
this
statement
in
the
whole
experience
is
the
male
28°
imaginal
series.
The
males
in
the
28°/28°
series
gave
an
abnormally
high
mean
duration
of
life.
It
is
of
some
interest
to
compare
the
relative
changes
in
body
size
and
in
duration
of
life
produced
by
different
temperatures
during
development.
From
Tables
3
and
4,
taking
the
four
most
reliable
measures
of
body
size,
femur
and
tibia
lengths
and
wing
length
and
breadth,
62
THE
AMERICAN
NATURALIST
[Vol—
LXIII
/000
/00
/0
j.
's
N
Devata",_
.
..
Oevat
/EV
16'
,
fria/s
Imaginal
Temperature
0
/5
.30
45
60
75
.90
/OS
/20
/35
/50
.19
,
e
in
days
FIG.
11.
Survivorship
lines
of
females
kept
during
imaginal
life
at
13°
C.
/05
iso
we
derive
the
following
percentages
regarding
reduction
in
body
size.
Character
Percentage
which
difference
between
1S°
and
28°
series
is
of
18°
flies
Males
Females
4
(femur
length)
7.0
5.7
5
(tibia
length)
6.9
6.0
8
(wing
length)
13.4
10.0
9
(wing
breadth)
13.1
9.4
Mean
of
four
characters
10.1
7.8
•No.
684]
THE
DURATION
OF
LIFE
63
/000
Datalir
/00
Females
/rnaginal
Temperature
25°
/0
O
/5'
0
vs
60
75
90
Age
in
days
FIG.
12.
Survivorship
lines
of
females
kept
luring
imaginal
life
at
25°
C.
It
is
evident
that
wings
are
proportionately
more
re-
duced
in
size
than
are
the
legs
in
high
temperature
as
compared
with
low
temperature
fl
ies.
Furthermore,
the
reduction
in
size
as
a
whole
(average
of
four
measure-
ments),
as
well
as
in
each
separate
character
of
the
four,
between
18°
and
28°
fl
ies,
is
greater
in
males
than
in
females.
64
THE
AMERICAN
NATURALIST
[Vol,.
LICHT
/000
DeVat/It°
100
/0
females
407idginal
Temperature
;
2g°
1
0
15
30
45
60
75
90
Age
in
days
FIG.
13.
Survivorship
lines
of
females
kept
during
imaginal
life
at
28°
C.
Corresponding
percentages
for
duration
of
life
of
fl
ies
reared
at
18°
and
28°,
respectively,
are
as
follows
:
Temperature
during
Percentage
which
difference
between
flies
imaginal
life
reared
at
18°
and
28°
is
of
18°
flies
Males
Females
19.5
7.6
25°
......
.........
16.1
12.7
-2
,
.
7.0
Mean
of
all
series...
17.8
9.1
From
these
fi
gures
it
is
seen,
fi
rst,
that
the
reduction
in
mean
duration
of
life
resulting
from
rearing
the
eggs,
larvae
and
pupae
at
28°
as
compared
with
18°,
is
some-
No.
684]
THE
DURATION
OF
LIFE
65
what
greater
proportionally
than
the
corresponding
re-
duction
in
body
size,
but
still
the
two
sets
of
fi
gures
are
of
the
same
general
order
of
magnitude.
Furthermore,
it
is
apparent
that
proportionate
reduction
in
duration
of
life
as
a
result
of
high
temperature
is
greater
in
males
than
in
females,
just
as
it
is
in
body
size.
These
results
indicate
with
a
considerable
degree
of
probability
that
in
these
experiments
the
quantitative
effects
of
temperature
differences
upon
the
biological
processes
concerned
in
growth
are
of
approximately
the
same
order
of
magnitude
as
the
quantitative
effects
of
temperature
differences
upon
the
biological
processes
concerned
in
the
determination
of
duration
of
imaginal
life.
This
is
the
kind
of
numerical
result
which
-
would
be
expected
on
the
rate
of
living
theory
of
life
duration,
because
in
both
cases
the
effect
of
increased
temperature
is
to
speed
up
the
rate
of
the
biological
processes
in-
volved.
In
the
18°
flies
we
have
a
slow
rate
of
energy
expenditure
in
growth
and
during
imaginal
life
(flies
very
inactive),
and
we
should
therefore
expect
on
the
theory
the
lengthened
duration
of
imaginal
life
which
we
observe.
In
the
28°
fl
ies
there
is
a
short
develop-
mental
period
and
a
consequent
rapid
rate
of
energy
expenditure
during
growth,
and
during
imaginal
life
(flies
very
active).
This
leads
to
the
expectation
of
a
short
duration
of
imaginal
life,
which
is
in
fact
observed.
VIII
Summarizing,
it
has
been
shown
in
this
paper
that:
1.
Prosophi/a
reared
at
18°
C.
is
distinctly
larger
in
a
series
of
bodily
dimensions
than
when
reared
at
28°
C.
Furthermore,
the
pigmentation
is
different
in the
two
cases,
notably
with
reference
to
the
pigmentation
of
the
seventh
tergum.
2.
Females
were
longer
lived
than
males
in
all
series
of
these
experiments.
The
difference
in
duration
of
life
between
the
two
sexes
diminishes
as
the
temperature
66
THE
AMERICAN
NATURALIST
[Vol,.
LXTII
during
imaginal
life
increases,
in
these
experiments,
re-
gardless
of
the
temperature
during
development.
3.
As
the
temperature
during
imaginal
life
increases
the
duration
of
life
decreases.
The
relationship
be-
tween
duration
of
life
and
temperature
is
exponential.
4.
For
the
temperature
range
18°
to
28°
used
in
these
experiments
the
temperature
coefficient
for
duration
of
life
had,
as
average
fi
gures,
the
following
values:
2.07
=
12717.3
5.
The
relative
or
proportional
influence
of
tempera-
ture
upon
body
size
and
upon
duration
of
life
was,
in
these
experiments,
of
the
same
order
of
magnitude.
This
fact
furnishes
confirmatory
evidence
to
the
theory
that
an
important
factor
in
determining
the
duration
of
life
is
the
rate
of
energy
expenditure
during
life.
LITERATURE
CITED
(The
plan
of
numbering
citations
is
explained
in
the
second
of
these
studies,
AMERICAN
NATURALIST,
Vol.
56,
p.
174.)
94.
Bliss,
Chester
J.
"Temperature
characteristics
for
prepupal
develop-
ment
in
Drosophila
inelanoyaster."
Jour.
Gen.
Physiol.,
Vol.
9,
pp.
467-495,
1926.
95.
Bonnier,
Gert.
"
Temperature
and
time
of
development
of
the
two
sexes
in
Drosophila."
British
Jour.
Exper.
Biol.,
Vol.
4,
pp.
186-
195,
1926.
96.
ll
iber,
R.
"Physikalische
Chemie
der
Zelie
und
der
Gewebe."
Gtr
Aufl.
Leipzig
(Engelmann),
192(1.
Pp.
xvi
+955.
97.
Janisch,
E.
"'Mier
die
Temperaturabhiingigkeit
biologischer
Vor-
giinge
und
ihre
kurvenmitssige
Analyse."
Arch.
f.
d.
yes.
Phys.,
209,
pp.
414-435,
1925.
98.
Id.
"Das
Exponentialgesetz
als
Grundlage
einer
vergleichenden
Biologie."
Abh.
zur
Theorie
der
organischen
Entwicklung,
Heft
II,
pp.
1-383,
1927.
99.
Krogh,
A.
"On
the
influence
of
temperature
on
the
rate
of
embryonic
development."
Zeitschr.
f.
ally.
Phys.,
Bd.
16,
pp.
163-177,
1914.
100.
Loeb,
J. "flier
den
Temperatuilioefficient
ffir
die
Lebensdauer
kaltbliitigei
Time
und
fiber
die
Ursache
des
naturlichen
Tories."
Arch.
f.
yes.
Plos.,
124,
pp.
411-426,
1908.
101.
MacArthur,
J.
W.,
and
W.
II.
T.
Baillie.
"Sex
differences
in
mor-
tality
and
metabolic
activity
in
Daphnia
magna."
Science,
Vol.
64,
pp.
229-230,
1926.
No.
684]
THE
DURATION
OF
LIFE
67
102.
Peak's,
L.
M.
"The
relation
of
temperature
to
insect
development."
Jonr.
Econ.
Entomology,
Vol.
7,
pp.
174-179,
1914.
103.
Pearl,
R.
"The
Rate
of
Living."
New
York
(Alfred
A.
Knopf),
1928.
Pp.
vii
+
185.
104.
Id.
"Experiments
on
longevity."
(1-z/art.
Review
of
Biology,
Vol.
3,
pp.
391-407,
1928.
105.
Titschack,
E.
"Untersuchungen
iiber
den
Temperatureinfluss
auf
die
Kleidermotte."
Zeitschr.
f.
wdss.
Zool.,
Bd.
124,
p.
213,
1925.
106.
Id.
"'Ober
die
huaginale
Lebensdauer
der
Kleidermotte."
Verh.
d.
Naturhist.
Vereins
Preitss.
2c.
Westf.,
Bd.
82,
p.
330,
1926.
107.
Wheeler,
W.
M.
"The
Social
Insects;
Their
Origin
and
Evolution."
New
York
(Harcourt,
Brace
and
Co.),
1928.
Pp.
xviii
+378.