re Ligh OF v 3 : SA
8 Ated tbe bak £68 R OM ; ; AS ress
AGE edie tae EN $ ; >

Crary

ae . ; 5
s > a3 Sos.

he
Kcthaticabe pees

eek ey

PRIA ee
SI ee ets 6
Pw ears,

~ oa 4 :
Spe thenact tn MS faet bali

Ppt iT ‘ FIG
ee, Coe eH

For OY weber
LS

I; ro

EN
1 AS
Rope er
Pe ee

(URAL HISTORY

i

SURVEY

aN aii Mary . :
as he i
a
a
| | a’,

ld

eA 7) ;
9
hay in oe 7 ZT

uF "

Hh Tyas
"

\

7

STATE OF ILLINOIS
DEPARTMENT OF REGISTRATION AND EDUCATION
NATURAL HISTORY SURVEY DIVISION

SIGNIFICANCE OF PESTICIDE RESIDUES:
PRACTICAL FACTORS IN PERSISTENCE

George C. Decker

Illinois Natural History Survey

Biological Notes No. 56
Urbana, Illinois March, 1966

WATURAL HISTORY Survey
APR }4 1966

UBRARY

SIGNIFICANCE OF PESTICIDE RESIDUES:
PRACTICAL FACTORS IN PERSISTENCE

OBVIOUSLY ALL FACTORS THAT INFLUENCE
the magnitude of pesticide deposits or the persistence of
residues have some practical value or significance, and
one finds it difficult to classify the factors involved as
either practical or theoretical Here the question of
practicality is really one of rationality—in other words,
correlation of the soundness of the evaluations we have
placed on every one of the known factors with the eff-
ciency with which we have utilized the available knowl-
edge.

Until quite recently the insecticides in common use
could be divided into groups, such as the highly volatile
fumigants, the rather unstable botanicals, and the stable
and persistent metallic salts. Thus it developed that for
many years most residue problems were associated with
highly stable and practically nonvolatile mineral salts.
Once such materials were applied to plants or other sur-
faces, they might be expected to persist unchanged for
indefinite periods of time or until removed by mechanical
processes.

FACTORS AFFECTING RESIDUE LOSSES

From time to time investigators working on residue
problems have isolated and evaluated some of the vari-
ous factors that affect the magnitude of residues and their
persistence. Most of these factors are fully discussed
and summarized in recent text or reference books (Brown
1951; Frear 1942; Shepard 1951) and therefore will not
be reviewed here. It is possible, however, that one or
more important factors involved have been largely over-
looked or grossly underestimated. The terms ‘vapor
pressure’ and “evaporation” are seldom given much
space, and in many cases are not even mentioned in
residue discussions. Perhaps this was fitting and proper
so long as we were primarily concerned with the residues
of lead arsenate and similar, essentially nonvolatile com-
pounds. We may have erred, however, in following the

This was an invitation paper, presented in Milwaukee, Wis-
consin, in April, 1952, as part of a symposium sponsored by the
American Chemical Society. It was accepted for publication
in a contemplated number of the Advances in Chemistry series,
but after this and several other manuscripts had gone as far as
the galley proof stage, the whole project was abandoned and the
series was never published. Meanwhile, Gunther & Blinn and
others had received the manuscript and had cited it as “In
Press.’ In response to the many inquiries as to the status and
disposition of the manuscript, to clear the record it is being
published here. Despite the fact that much progress has_ been
made in the intervening years, in fairness to those who read and
commented upon the original manuscript, it is published almost
unaltered and exactly the same as it would have appeared had
publication been consummated in 1953.

This paper is published by authority of the State of Lllinois,
IRS Ch. 127, Par. 58.21. It is a contribution from the Section
of Economic Entomology of the Illinois Natural History Survey.

Dr. George C. Decker, formerly Principal Scientist and Head
of the Section of Economic Entomology, Illinois Natural History
Survey, is now retired.

GEORGE C. DECKER

same old lines of approach when we began to attack
problems associated with the use of the chlorinated hy-
drocarbons and other new, synthetic, organic insecticides.

Very soon after DDT came under intensive study in
1944, Fleck (1944:853), reporting on the “Rate of Evapo-
ration of DDT,” concluded that “the loss of DDT from
insecticidal spray deposits by volatilization will occur
too slowly to be of any importance.” Two years later
Wichmann et al. (1946:218-233) apparently came to
about the same conclusion but did not clearly say so.
Unfortunately, in both instances the rate of evaporation
was determined by exposing known amounts of DDT
crystals (63.36 mg and 200 mg) on glass plates which
were weighed at intervals. This method, of course, was
not conducive to the production of maximum losses by
evaporation.

About the same time, Gunther et al. (1946:624-627),
reporting on rather extensive residue studies involving
progressive analyses of apple foliage and fruit samples,
showed that after 86 days “every treatment showed a
loss of from 71 to 95 per cent of the original quantity
of DDT deposited.” There was no suggestion or impli-
cation, however, that even a part of the loss might have
been due to evaporation. The final conclusion was
merely, “A distinction between mechanical weathering
and chemical decomposition has not been attempted in
this report.”

Fleck (1948:706—708) introduced a noteworthy paper
with the following appropriate and highly significant
statement, ‘The residual action of an insecticide is de-
termined by its vapor pressure, its sticking power, its
solubility, its absorption into the surface to which it is
applied, and its resistance to chemical change.” From
that point on, however, the theme of the paper is chemi-
cal change or decomposition; vapor pressure and evapo-
ration are ignored.

It would appear that most workers have taken their
cue from these and similar reports, for although several
writers (Hadaway & Barlow 1951:854; Hensill & Gard-
ner 1950:102-107; Walker 1950:123-127) have made
some reference to vapor pressure, volatility, or evapora-
tion, no one has come forward to emphasize the impor-
tance of evaporation as a factor strongly influencing the
rate of residue loss where the chlorinated hydrocarbon
insecticides and similar materials are involved. Even
where a significant and consistent progressive loss of
residue has been noted, the tendency has been to at-
tribute the losses almost entirely to erosion, weather, or
chemical decomposition.

More recently, in 1950, the writer and his associates
(Decker, Weinman, & Bann 1950:919-927), in reporting

PER CENT OF INITIAL DEPOSIT

DAYS AFTER SPRAYING

100

PER CENT OF INITIAL DEPOSIT

| 2 a 7 14
DAYS AFTER SPRAYING

TABLE 1—Vapor pressure at 25°C. and per cent loss of
residue from apple leaves in 10 days for five chlorinated hydro-
carbon insecticides.

Per Cent Residue

Vapor Pressure

Insecticide atoonG; Loss in 10 days
Lindane 0.000,010 95
Aldrin 0.000,052 89
Chlordane 0.000,010 80
Dieldrin 0.000,002,5 65
DDT 0.000,000,3 46

residue studies involving apples, peaches, soybeans, al-
falfa, and clover, showed that under varied conditions
of exposure the residues of lindane, aldrin, chlordane,
dieldrin, toxaphene, and DDT tend to disappear in the
order named, and usually in about the same relationship.
It was shown also that in some cases, at least, the per
cent of the initial deposit remaining at intervals was a
straight-line logarithmic function of time (Fig. 1 and
2). Evaporation was not clearly established as an impor-
tant factor in residue loss, but data were presented to

Fig. 1—Average per cent
of initial residue remaining on
apple leaves at intervals after
spraying with several insecti-
cides.

Fig. 2.—Average per cent
of initial residue remaining at
intervals after red clover was
sprayed with several insecti-
cides.

show that there was “a considerable loss of even slowly
volatile materials during the actual spraying process.”
Later the apparent correlation between the reported
vapor pressures of the compounds in question and the
order in which their residues tended to disappear (Table
1) led to the conclusion that volatility might be an im-
portant factor in determining residue persistence.

EVAPORATION STUDIES

In the light of the above observations the writer and
his associates set about to explore the ramifications of
and, if possible, determine the probable importance of
this relatively unevaluated factor.

Throughout the investigations reported, the chemical
analysis procedure utilized was that devised by Stepanow
(1906:4056-4057) with the modification suggested by
Fleck (1947:319-324), Umhoefer (1943:383), and Cald-
well & Moyer (1935:38-39). The method involves the
conversion of organic chlorine to the chloride ion, which
is subsequently determined by the Volhard titration

4

method. The sensitivity of this method is generally
considered to be in the neighborhood of 1.0 ppm, though,
in the vast majority of cases, replicates were in close
agreement down to 0.5 and 0.3 ppm. The use of this
method, however, leaves open to speculation the proba-
bility that the trends shown may continue below the
1.0 ppm level.

In the process of developing techniques for the va-
porization study—which began with the exposure of
known quantities of material in petri dishes or on glass
slides and terminated in the use of 10 to 50 mg deposits
of crystalline material uniformly dispersed over 20 by 20
inch sheets of semicrepe, white filter paper—it was shown
that with a large mass of material heaped or piled on a
small area the per cent loss in weight at various intervals
was very small, but as the mass of material per unit of
surface area was decreased, the rate of loss increased.
Finally it was found that the residues deposited on filter
paper responded to external conditions much the same
as did normal residues on plants (Table 2).

TABLE 2.—Residue loss in 14 days’ exposure under green-
house conditions (temperature 80 + 5°F., R.H. 75 = 5 per cent)
for seven insecticides.

Per Cent Loss on
20 by 20 Inch

Per Cent Loss on Glass Plates Filter Paper
Insecticide 100-150 mg 10-25 mg 10-50 mg
Parathion 96 — 98.6
Lindane 82 — 95.4
Aldrin 63 90 90.6
Chlordane 50 88 82.0
Dieldrin 41 52 68.0
Toxaphene 24 31 49.0
DDT 21 26 44.0

Under controlled conditions in the laboratory, where
many of the variables could be eliminated, the writer
and his associates (Decker, Weinman, & Bann 1950:
919-927) found that the data obtained were exceedingly
consistent and in most cases, when plotted, all points
were on, or in close proximity to, the lines mathematt-
cally fitted to the data. Under such conditions, residue
losses for eight and, in some instances, more insecticides
were compared at five temperatures. In all experiments
the order of loss was the same, with lindane disappear-
ing first and DDT last. As was previously reported
(Caldwell and Moyer 1935:38-39) for the field experi-
ments, the residue-loss curves for pure compounds
showed that the loss was a straight-line logarithmic func-
tion of time (Fig. 3). In general, the data for impure
substances, such as technical aldrin, chlordane, and toxa-
phene, when plotted, departed from the straight line
and produced typical wavy lines characteristic of mixtures
of compounds differing in their volatility. This was
found to be true both on foliage in the field and under
controlled conditions in the laboratory. This tendency
is evident in Fig. 1 and 2.

Changes in temperature, of course, produce corre-
sponding changes in the vapor pressure of each substance,

and thereby alter the rate of evaporation or residue loss.
The work in progress at Illinois showed definitely that,
although changes in temperature affected rates of evapo-
ration or residue loss, in general they produced some-
what corresponding effects on all of the materials studied,
as shown for aldrin and DDT in Fig. 4. While an in-
crease in wind or air movement tends to accelerate the
rate of residue loss, it affects all materials more or less
alike and does not greatly alter their relative positions.

A review of all the conflicting data obtainable led to
the obvious conclusion that several factors influence the
rate of evaporation and that vapor pressure alone is not
an accurate measure of that loss. This is particularly
true for the initial period of each test when the spray
is being applied and drying on the surface; and it is
evident throughout some tests. Evaporation, therefore,
must be considered as the summary effect of the various
factors which influence residue loss through vaporiza-
tion: vapor pressure (probably the most important),
ratio of mass to surface exposed, air movement, type of
formulation used, etc. It follows logically, then, that
a substance like DDT, with a vapor pressure of approxi-
mately 3.3 x 10% at 25° C., exposed as a mass in an open
container would lose weight very slowly and might be
considered practically nonvolatile. As a thin layer of
fine, fluffy crystals exposed to warm, moving air, the
rate of loss would be increased manyfold, and it might
then be regarded as fairly volatile.

Presumably evaporation has been established as an
important, if not a dominant, factor influencing residue
persistence for many of our presently used and poten-
tially available insecticides. That brings us back to
Fleck’s highly significant statement (1948:706—708),
“The residual action of an insecticide is determined by
its vapor pressure, its sticking power, its solubility, its
absorption into the surface to which it is applied, and
its resistance to chemical change.”

In most instances one should know or determine in
advance the relative importance of such factors as solu-
bility, resistance to chemical change (stability), and
probability of absorption into plant tissue or coatings.
Such factors can then be properly evaluated or elimi-
nated from further consideration. That leaves vapor
pressure and sticking power to be considered, and to
these I would add the probability of dilution owing to
plant growth and formulation variables. While there
would, of course, be exceptions to the rule, it is quite
probable that the last three factors would apply about
equally or proportionately to all pesticide deposits that
may come under consideration. That is, regardless of
the pesticide used, the factor of plant growth would
diminish or dilute all equally. Likewise, in the case of
sticking properties and formulation variables, it seems
probable that the methods of formulation and applica-
tion employed would determine tenacity and would
apply equally to all materials without regard to their
toxicity or volatility. If three of the four big factors—
tenacity, dilution by plant growth, and differences at-

100

804
b
”
co}
a
Ww
oc 60
aa
c4
‘=
2 DIELDRIN
L
° 40
e
a ALDRIN
o
c
w
a 20

! 2 4 8 24 48
TIME IN HOURS

PER CENT OF INITIAL DEPOSIT

Fig. 3.—Average per cent
of initial residue remaining at
intervals after 20 by 20 inch
{lter papers bearing 10 mg of
several insecticides were placed
in a constant temperature
chamber at 80°F.

—
96 144 1,000

Fig. 4.—Average per cent
of initial residue remaining at
intervals where 10 mg de-
posits of aldrin and DDT were
exposed on filter papers to
varied temperatures.

| 10 100
TIME IN HOURS

tributable to formulation—apply about equally to all
residues, vapor pressure remains the one important vari-
able which might largely determine relative persistence
of the residues produced by various pesticidal chemicals.

PRACTICAL APPLICATION OF GENERAL PRINCIPLES

What are the practical implications or considerations
involved? Presumably we are interested in residues pri-
marily for their public health aspects. Therefore, our
first concern should be how we may most efficiently util-
ize chemicals to control pests and at the same time assure
the public the greatest possible degree of safety. Where
we have a choice of pesticides that will do a given job,
we must decide in each instance which of the materials
available will be most effective and which can be used
with the greatest degree of safety.

Obviously, the inherent toxicity of the chemicals in
question to man and other warm-blooded animals must
be taken into account, but it may be a minor factor and
by no means the dominant consideration. Very often

1,000

1
10,006

100,000

one may find that, from the standpoint of residues, the
most toxic substance will be the safest one to use by a
considerable margin. In most instances, the more toxic
chemicals are used or applied in proportionately smaller
amounts than are the less toxic materials, and frequently,
but not always, the more toxic compounds are the most
volatile and are therefore short-lived.

On the basis of data presented and of many fragments
of data obtainable elsewhere, one must conclude that
the rate of loss through evaporation for an insecticide is
basically a function characteristic of that compound (an
index combining vapor pressure, size, and shape of crys-
tal formation, etc.). Through all the work in field and
laboratory it was noted that the percentage of residue
loss from day to day and from week to week was re-
markably consistent for each compound and quite or
completely independent of the dosage rate or magnitude
of the initial deposit. Thus, it appeared that insofar as
loss through evaporation is concerned, residues resulting
from different rates of application of a given substance

40
w
Kb
: S0
& 30 MG
a
gq
2 QuG
Fad
2 20F
aq 30
oO MG e
=
e
e
2
1ob OMG =
Neal = _
| 2 4 8 24 48

TIME IN HOURS

will, under the same conditions of exposure, reach the
vanishing or zero point at exactly the same time regard-
less of the magnitude of the original deposits. This
sounded logical, but left a feeling of apprehension. In
all instances, however, when this theory was tested by
exposing various insecticides at varying rates of appli-
cation, the assumption proved to be valid. All deposits
of each compound similarly exposed reached the vanish-
ing point at the same time (Fig. 5).

Now, let us see what the general principles just de-
veloped mean in terms of practice. In the first place,
with several materials available, what would be the pos-
sibility or probability that if used to control a certain
pest one or more of the materials would leave a detect-
able residue at harvest time? Where adequate experi-
mental data are available, the answer to that question is
fairly simple. Since it has been shown that under any
given set of conditions the residues produced by a given
insecticide will arrive at the vanishing point or zero
level at a specific time, regardless of the rate of applica-
tion or the magnitude of the initial deposit, a very good
indication of which materials will be most likely to show
residues at harvest time can be obtained from Fig. 1, 2,
and 3. They indicate the time required for the residue
of each material studied to reach the base line or zero
point. Obviously, lindane would have the best chance
of showing no residue, followed in order by aldrin,
chlordane, dieldrin, toxaphene, and DDT, with little
likelihood that a DDT residue will ever reach the van-
ishing point unless aided very materially by other im-
portant factors, such as plant growth or erosion. Some
idea of the odds that residues of the various materials
would be gone by harvest, or by any other given time,
might be obtained by comparing the data on days re-
quired to reach zero (Table 3).

From these data, which may be subject to consider-
able error, one may get a fair picture of the relative
importance and apparent practical value of several fac-
tors that affect the rate of residue loss. From the labo-

96 144

Fig. 5—Aldrin residues in
milligrams at intervals after
20, 30, 40, and 50 mg de-
posits on 20 by 20 inch filter
papers were placed in 80°F.
temperature chambers.

1000

_ TABLE 3.—Days of exposure required for residues of various
insecticides to decline to zero point under field and laboratory
conditions.

Approximate Days Required to Reach Zero

Lab. Papers
Insecticide Fruit Foliage Clover Plants at 80°F.
Lindane 15 — 35
Aldrin 22 15 EA
Technical aldrin 27 — 9.2
Chlordane 35 21 185.0
Dieldrin 45 25 2,200.0
Toxaphene 85 40 140,000.0
DDT 105 50 210,000.0

ratory data one obtains a forceful impression of the dif-
ference in evaporation rates of the semivolatile materials,
such as lindane and aldrin, in contrast to the relatively
nonvolatile materials, such as toxaphene and DDT. The
importance of evaporation in the case of all materials
including dieldrin is evident, but a study of the values
for toxaphene and DDT discloses the very great signifi-
cance of erosion and weathering where mature foliage
was involved, and the added effect of plant growth in
diluting the residue in the case of the growing clover
plants.

APPLICATION OF MATHEMATICAL PRINCIPLES

Assuming that the interval between the date of last
treatment and harvest will be such that none of the
materials, not even lindane, will have time to vanish
completely, the probable magnitude of the residue of
any individual compound will be in direct proportion
to the dosage rate or the magnitude of the original de-
posit. As has been shown in the comparison of several
dosage rates, since the per cent of original deposit re-
maining at the end of each interval of time is the same,
the residues at any specified interval will be in direct
proportion to the magnitude of the original deposit.
For example, in Fig. 5, where the original deposits were
in the approximate ratios of 2, 3, 4, and 5, the residues
remaining in each case on the Ist, 2nd, 4th, and 6th

days were still in those same relative ratios. In addition,
the ratios were maintained to the point at which, insofar
as the graph is concerned or mathematical calculations
can determine, some time on the 12th day (300 hours),
according to both theory and practice, all residues dis-
appeared simultaneously.

If one wished to compare two materials, he would
have to take into account the slope of the line which
can be established by use of the formula Y — a + bX
(where X is the logarithm of time; b, the regression co-
efficient; and a, a constant derived from the equation,
a — Y — bX), and the normally recommended or
probable rates of application. The higher the evapora-
tion rate and the lower the normally recommended rate
of application, the greater the probability that the residue
will disappear or reach an insignificantly low level in
X days or any other given period. Conversely, the lower
the evaporation rate and the higher the dosage or appli-
cation rate, the greater will be the probability that a
very significant residue will be present after any given
period.

It is to be hoped that further study of the mathe-
matical principles involved may lead to the development
of a formula or formulae which will make it possible
to compare two or more insecticides and predict rather
precisely the probable relative magnitude of their re-
spective residues X days after treatment, or the relative
number of days required for each to reach a residue of
Y magnitude. If this becomes a reality, then it will be
possible to multiply the values obtained by some suitable
index of chronic toxicity supplied by the toxicologist
to obtain in advance a fair estimate of the probability
that a specified use of a pesticide would result in a sig-
nificant food contamination hazard. Until such time as
suitable formulae are developed, helpful comparisons

560

TOXAPHENE

400

RESIDUE IN PPM ON APPLE LEAVES

120
—_LINp,
A A
80 LORIN O25 es

0.25 LB,

DAYS AFTER TREATMENT

7

may be made by utilizing data on dosage rates and 50 per
cent or 10 per cent life values.

It is possible also that graphic presentations, such
as Fig. 6, will prove helpful in visualization of some of
the complex dosage-rate vs. residue-loss relationships.
For example, in comparing lindane or aldrin with the
reportedly less toxic materials, DDT or toxaphene, one
finds that on the 15th day the residues in ppm are, rough-
ly, lindane, 1.0; aldrin, 10; DDT, 152; and toxaphene,
220. On the 20th day they are lindane, 0; aldrin, 2;
DDT, 130; and toxaphene, 185. It would appear that
in estimating ultimate safety, one would have to balance
these or other comparable values for any given time
against the relative toxicities of the materials in question.
While the ppm lines (Fig. 6) are admittedly calculated
from hypothetical considerations, a check on seven points
where good data are available indicates all lines are well
within a 10 per cent error.

DISCUSSION

The data and discussion presented here in no way
establish vaporization as the sole or even the dominant
factor in determining the rate of residue losses. We are
all too familiar with the importance of weathering, dilu-
tion due to plant growth, etc., to discount the impor-
tance of these factors. In the field phases of investiga-
tions reported here it was impossible to isolate such
factors and study them separately. The losses observed
in the field investigations were in reality the total ac-
cumulative effects of all factors combined. In the final
analysis, however, it seems significant that in comparing
the residue losses of the different materials tested they
tended to persist in the inverse order of their vapor
pressures. This could only mean that with the added

100

80 Fig. 6.—Hypothetical res.
idues in parts per million
and per cent of initial resi-
due at intervals after treat-
ment, assuming four mate-
rials were sprayed on apple
foliage at normally recom-
mended rates of application.

PER CENT OF INITIAL DEPOSIT

8

advantage of a relatively high vapor pressure, the residue
of a given insecticide would tend to disappear more rap-
idly than the residue of another insecticide lacking high
vapor pressure.

Time will not permit a lengthy discussion of other
problems, but perhaps the enumeration of a few other
practical considerations worthy of thought may be in
order.

(1) Where we have been making progressive resi-
due analyses under field conditions, we have found that
in some instances after a residue had shown a typical
straight-line decline to near zero, we suddenly had a
very great rise in the residue level (sometimes 20 to 100
ppm). This was inevitably traceable to drift from other
spray operations in the vicinity, in some cases up to 750
feet distant. This raises the question of how much
confidence we can place on harvest residues alone unless
we are sure beyond doubt that no contamination oc-
curred between our recorded treatment and harvest.

(2) It was observed that where there is any air
movement at all, the magnitude of an initial deposit
increases row by row from the windward margin of a
plot and that there is a very appreciable deposit several
rows beyond the plot. One may therefore question the
validity of small-plot data.

(3) Since the solvents employed will influence the
time of crystallization and the type of crystals produced,
to what extent have formulation differences clouded our
results and to what extent can we utilize formulation
differences to hasten or retard the disappearance of a
residue?

LITERATURE CITED

BROWN, A. W. A. 1951. Insect control by chemicals. John Wiley
& Sons, New York.

CALDWELL, JOHN R., AND HARVEY V. MOYER. 1935. Determi-
nation of chloride. Industr. and Engin. Chem., Analyt. Ed.
ME) 38—39:

DECKER, G. C., C. J. WEINMAN, AND J. M. BANN. 1950. A
preliminary report on the rate of insecticide residue loss from
treated plants. Jour. Econ. Entomol. 43(6) : 919-927.

FLECK, ELMER E. 1944. Rate of evaporation of DDT. Jour.
Econ. Entomol. 37(6): 853.

. 1947. Report on methods for analysis of DDT and
insecticidal preparations containing DDT. Assoc. Off. Agr.
Chem. Jour. 30(2) : 319-324.

. 1948. Residual action of organic insecticides. In-
dustr. and Engin. Chem. 40(4) : 706-708.

FREAR, DONALD E. H. 1942. Chemistry of insecticides and fungi-
cides. D. Van Nostrand Co., Inc., New York.

GUNTHER, F. A., D. L. LINDGREN, M. I. ELLIOT, AND J. P.
LADUE. 1946. Persistence of certain DDT deposits under field
conditions. Jour. Econ. Entomol. 39(5): 624-627.

HADAWAY, A. B., AND F. BARLOW. 1951. Sorption of solid
insecticides by dried mud. Nature 167 (4256): 854.

HENSILL, G. S., AND L. R. GARDNER. 1950. Some poisonous
residue factors in use of two new organic insecticides. Ad-
vances in Chemistry Ser. 1:102—-107.

SHEPARD, HAROLD H. 1951. The chemistry and action of in-
secticides. McGraw-Hill Book Co., New York.

STEPANOW, A. 1906. Ueber die Halogenbestimmung in or-
ganischen Verbindungen mittels metallischen Natriums und
Aethylalkohol. Ber. 39:4056-4057. Med. Chem. Lab., Univ.
of Moscow. (Chem. Abs. 1:397).

UMHOEFER, ROBERT R. 1943. Determination of halogens in
organic compounds. Industr. and Engin. Chem., Analyt. Ed.
15 (6) : 383-384.

WALKER, KENNETH C.
fruits, apples, and pears.
We P27/e

WICHMANN, H. J., W. I. PATTERSON, P. A. CLIFFORD, A. K.
KLEIN, AND H. V. CLABORN. 1946. Decomposition and
volatility of DDT and some of its derivatives. Assoc. Off.
Agr. Chem. Jour. 29 (2) : 218-233.

1950. Parathion spray residue on soft
Advances in Chemistry Ser. 1:

(25917—1M—3-66)
offices 14

_

SPA