DATE DUE

-

UNIVERSITY OF MASSACHUSETTS
LIBRARY

s

73
E5

no.U3
1903-07

TECHNICAL BULLETIN.

HATCH EXPERIMENT STATION

OF THE

MASSACHUSETTS

AGRICULTURAL COLLEGE.

/vo. /.
THE GREENHOUSE ALEYRODES (A. vaporahorum Westw.)

AND THE

STRAWBERRY ALEYRODES (A. packardi Morrill):

A STUDY OF THE INSECTS AND OF THEIR TREATMENT.

By AUSTIN W. MORRILL, Ph. D.

AUGUST, 1908.

The regular Bulletins of this Station will be sent free to all news-
papers in the State and to such individuals interested in farming as
may request the same. Technical Bulletins are sent only to those persons
interested in the subject treated of i?i each ease.

AMHERST, MASS.:

PRESS OF CARPENTER & MOREHOUSE,
1903-

HATCH EXPERIMENT STATION

OF THE

Massachusetts Agricultural College,

AMHERST, MASS.

ST-A.TI03ST STAFF:

Henry H. Goodell, LL. D.,
William P. Brooks, Ph. D.,
George E. Stone, Ph. D.,
Charles A. Goessmann, Ph. D.,LL.D
Joseph B. Lindsey, Ph. D.,
Charles H. Fernald, Ph. D.,
Frank A. Waugh, M. S.,

J. E. OSTRANDER, C. E.,

Henry T. Fernald, Ph. D.',
Frederick R. Church, B. Sc.,
Neil F. Monahan, B. Sc,
Henri T). Haskins, B. Sc,
James E. Halligan, B. Sc.
Richard H. Robertson, 13. Sc,
Edward B. Holland, M. S.,
Philip H. Smith, B. Sc,
William E. Tottingham, B. Sc,
Albert Parsons, B. Sc,
Joseph G. Cook, B. Sc,

Fred. F. Henshaw,

Director.
Agriculturist.
Botanist.

,Chcmist (Fertilizers).
Chemist (Foods and Feeding).
Entotnologist.
Horticulturist.
Meteorologist.
A s socio te En torn ologist.
A ssistant Agriculturist.
A s sis la u t Bo to n ist.
First Assis't Chemist (Fertilizers).
Second Assist Chemist (Fertilizers).
Third Assist Chemist (Fertilizers).
First Chemist (Foods and Feeding).
AssH Chemist (Foods and Feeding).
AssH Chemist (Foods and Feeding).
Inspector (Foods and Feeding).
Assistant in Foods and Feeding.
A ssistant Horticulturist.
A ssista n t Horticu Itu rist.
Observer.

The co-operation and assistance of farmers, fruit-growers, horti-
culturists, and all interested, directly or indirectly, in agriculture,
are earnestly requested. The Bulletins will be sent free to all news-
papers in the State and to such individuals interested in farming as
may request the same. General bulletins, fertilizer analyses, analy-
ses of feed-stuffs, and annual reports are published. Kindly indi-
cate in application which of these are desired. Communications
may be addressed to the

Hatch Experiment Station, Amherst, Mass.

1 .?

• Ml

ijMIVF • IF

}0.

■

CONTENTS.

Introduction, . . ...

Systematic Position of Aleyrodes,

Methods of Study, ......

The Greenhouse Aleyrodes.

Description, .....

Notes on Life History and Habits,

Identity of the Insect, ....

Origin and Distribution, .

Food Plants, .....

Economic Importance, . . . .

Preventives, . : . . .

Remedies, ......

Experiments with Contact Insecticides,
Experiments with Fumigants,
Discussion of Results,
Recommendations, . . . .

Bibliography, .....

The Strawberry Aleyrodes.

Description, ......

Notes on Life History and Influence of Weather,
Economic Importance, . .

Treatment, .....

Preventives, .

Remedies, .....

Bibliography, . . . . .

Summary, ......

Explanation of Plates, . . . .

5
7
9

12

28

34
34
34
35
36
37
38
4i
47
49
52

53
55
56
57
58
58
61
62
65

ADDENDA AND CORRIGENDA.

Unfortunately it was impossible for me to correct the proof of
this bulletin, but I take this opportunity to call attention to the errors
I find in its completed form.

A. W. MORRILL.

Victoria, Texas, Aug. 15, 1903.

Page 8, line 1 from top, for Therefore the Hotnoptera read The Ho7>ioptcra.

Page 11, line 2 from top, for by read after.

Page 14, line 13 from bottom, for which appears to extend read which
appears to be part of a spiny surface that extends.

Page 16, line 16 from top, for on a line about read on a line with the bases
of the legs of its respective side of the body about.
Page 20, line 14 from top, for letter C read letter f.
Page 24, line 14 from top, for dark white read black.

Page 26, line 11 from lop, for common read small .

1 11 in
Page 28, line 18 from top, for LLL read L L L.

Page 29, line 10 from top, tor from plant to read from one plant to.

Page 30, line 10 from top, for the sense the read the sense of the.

Page 33, line 7 from bottom, for series read species.

Page 35, line 11 from top, tor generous read general.

Page 42, line 5 from bottom, for Buck's read Bush's.

Page 46, line 9 from bottom, add Plants : Tomatoes, badly infested.

Page 47, line 1 from top, for cheek read check.

Page 47, line 1 1 from bottom for these latter read Insecticides of this class.

Page 49, line 4 from top, for inexpensive read rather expensive.

Page 50, line 20 from top, for $1.50 read $1.00.

Page 53, line 10 from top, read on strawberries in Ohio, etc.

Page 60, experiments Nos. i-\ 1 lorgms. per cu. ft. read lbs. per 1000 cu.ft.

Between experiments 11 and 12 read EXPERIMENTS WITH POTAS-
SIUM CYANIDE.

Page 63, line 6 from bottom, for .1 gram read .01 gram.

INTRODUCTION.

The following paper is the result of studies begun more than
two years ago at the Entomological Laboratory of the Massachusetts
Agricultural College, as a consequence of complaints made to the
Hatch Experiment Station of serious injury caused by the Green-
house Aleyrodes (Aleyrodes vaporariorum), and the need of more sat-
isfactory methods of controlling the pest than those previously
known. Mr. J. B. Knight, who was at that time pursuing the grad-
uate course for the degree of Doctor of Philosophy at the college,
began a study of the different stages of the insect, but having given
up the work to accept a position in India, the study of this insect
was taken up by the present writer. Mr. Knight's description was
briefer than the one here given, and has been used only for com-
parison, none of his work having been incorporated in this paper.

Having discovered in the course of my studies that the Straw-
berry and Greenhouse Aleyrodes were distinct species, rather than the
same, as was generally supposed, I have extended the scope of the
work as previously outlined, to include a discussion of both. At
present, there are about sixty-five American species of the genus
Aleyrodes known.* Of these, the two species which form the sub-
ject of this paper, A. vaporariorum Westw., and A. packardi Morrill,
with the Orange Aleyrodes, A. citri Riley and Howard, are the only
ones which have thus far proved to be of much economic importance
in the United States. A. citri, which was first described from Flor-
ida, has been recorded in greenhouses as far north as Michigan,!
and Quaintance % states that it occurs generally in greenhouses ; its
attacks, however, are almost wholly confined to orange and lemon
trees. This species has never been recorded from Massachusetts.

I have followed the original spelling of Latreille in the use of

* Including twenty Californian species, the descriptions of which are unpublished at
the time of this writing (Psyche, vol. ix, p. 32S.)

t Davis, Miscellaneous Bull. (Special Bull. No. 2) Mich. Expr. Station, p. 24, (1S96.)
% Quaintance. U. S. Dept. Agric, Div. ent., Tech. ser. Bull. No. S, p. 22, (1900.)

2

the generic name of these insects, no typographical error being evi-
dent. Four common names for the members of this group are known
to the writer, viz.: Aleyrodes, white fly, mealy wing and snowy fly,
the latter name apparently being in use only in England. The com-
mon name, Aleyrodes, derived directly from the name of the genus,
will be used hereafter in this paper in preference to the others.

The bibliography of each species is numbered, and whenever I
have had occasion to mention any of those writings I have referred
to them by their number only. Other references are found in the
foot notes.

This paper represents a little more than half of a thesis pre-
sented to the faculty of the Massachusetts Agricultural College in
June, 1903, for the degree of Doctor of Philosophy. I wish to ac-
knowledge here my obligations to Professors C. H. and H. T. Fer-
nald, under whose able supervision and direction I have pursued my
graduate studies in the Entomological Laboratory of the college dur-
ing the past three years.

Systematic Position of Aleyrodes.

The two genera, Aleyrodes Latreille and Alenrodicus Douglas,
constitute the family of Homoptera* called Aleyrodidae. The latter
genus is distinguished from the former by the presence of a distal
and a basal branch to the median vein in both pairs of wings. The
insects of this family were for a long time classed with the Coccidae,
or scale insects, on account of the resemblances between the imma-
ture stages. The differences between the two groups of insects in
their metamorphosis and the form of the adults have been used as
the chief characters to distinguish the family Aleyrodidae from the
Coccidae, and the former is now given a systematic position between
the latter family and the Aphidae, or plant lice. The larval and
pupal stages of the Aleyrodidae are distinguished from those of the
Coccidae by an opening on the dorsum of the last abdominal seg-
ment known as the " vasiform orifice." The insects undergo a com-
plete metamorphosis, thus differing in a marked degree from all other
Homopterous insects, except the Coccidae, where the male alone un-
dergoes a complete metamorphosis. In both these families the
metamorphosis, though complete, is of a peculiar nature, and differs
in some respects from that of the true Metabola, or such insects with
complete metamorphosis as Lepidoptera (butterflies and moths),
Diptera (flies), and Coleoptera (beetles). Both males and females
of the Aleyrodidae have in the adult condition two pairs of well-
developed wings, while in the adult male Coccidae the hinder pair of
wings is represented by minute appendages called halteres. There
is much evidence to show that incomplete metamorphosis is the an-
cestral condition of insects, and therefore in general, insects with
complete metamorphosis represent a higher degree of specialization
and are younger than those which have retained the former condi-

* For the benefit of those who are not familiar with the classification of insects, it is
here explained that the Homoptera and Heteroptera are two sub-orders of the Hemiptera.
The former is distinguished from the latter by the form of the wings, which are of the same
thickness throughout and usually sloping at the sides of the body, and by the position of
the origin of the beak, which is at the hinder part of the lower side of the head. The Hom-
optera include such insects as plant lice, scale insects, cicadas and leaf hoppers. To the
Heteroptera belong the true bugs, such as the squash bug and chinch bug.

8

tion. Therefore, the Homoptera are generally believed to be more
primitive than the Heteroptera. The Aleyrodidoe and Coccidae
represent the highest degree of specialization among the Homoptera*
and even surpass the Heteroptera in that both have attained a com-
plete metamorphosis. At first glance it appears that the former
family, having a complete metamorphosis in both sexes, is a more
specialized form than the latter family, but this is more apparent
than real. It is in accordance with our belief in regard to the origin
of complete metamorphosis that the ancesters of the Coccidae in the
adult condition were winged in both sexes, and that the adult form
was the result of gradual changes in the immature insects. The
Coccid branch, therefore, was derived from the primitive Homoptera
with incomplete metamorphosis. In the course of time the younger
stages, through the action of natural selection, departed from the
original type in which they resembled the adults, as do the other
Hemimetabola (insects with incomplete metamorphosis) to-day, so
that they approached and finally attained a complete metamorphosis
in both sexes. It is not impossible that the degeneration of the
female began during the latter part of this process. If this view be
correct, the female Coccids have now attained a secondary incom-
plete metamorphosis. The male Coccids were evidently prevented
from losing their wings by natural selection in order that they might
continue to be active in fertilizing the females. Even the adult
males are, however, degenerated, having lost their mouth organs, and
are so delicately constituted as to live but a few days at the most.
The Aleyrodidas have gone through a series of changes similar to
those of the Coccidae up to the point where complete metamorphosis
was attained, and it is at that point that we find them to-day, although
the different species are specialized in various other directions. In
tracing the two lines of development backward, beginning with our
present forms, do they unite before they reach the primitive Homop-
terous branch, or is it simply a case of development along parallel
lines owing to the two forms living under similar conditions ? The
general resemblance of the two families seems to point to a union of
the two lines, although, as a rule, the immature stages of insects are
of uncertain value in the study of phylogeny.

t The Cicadidre attain a high degree of metamorphosis, but it can hardly be con-
sidered complete.

The resemblance of some of the members of the Homopterous
families Psyllidae and Aphidae to Aleyrodes is worthy of mention.
Many years ago, Signoret, who was a prominent French entomolo-
gist and an authority on the Coccidae and Aleyrodidae, classed with
the latter family the genus Spondyliaspis, which Maskell* states be-
longs without doubt to the family Psyllidae. Among the Aphidae,
some of the Pemphiginae bear a striking resemblance to the larva
and pupae of Aleyrodes. These cases at least illustrate how insects
like the Aleyrodidae and Coccidae could have arisen from primitive
Homopterous forms which had well developed legs in all larval
stages, and well developed legs and two pairs of wings in both sexes
of the adult.

From our present knowledge it seems probable that the Aley-
rodidae and Coccidae are more closely related to one another than
either is to any other group, but further study of the least specialized
members of each family is necessary before any definite conclusions
can be drawn.

Methods of Study.

Aleyrodes, on account of their delicacy and small size, offer
some difficulties to a beginner. The immature stages are best
handled with a needle, aided by a good hand lens. The needle
should be inserted between the body of the insect and the leaf, pre-
ferably from the side. The insects can then be lifted and deposited
on a slide by rolling the needle, or if the larva is in the first instar
and crawling, it may be allowed to crawl off the needle point. In
the latter case the slide should then be placed in a strong cyanide
bottle for about five minutes. No cover glass need be used for ordi-
nary study with a one-fifth or even Zeiss F objective. To use an oil
immersion lens a thin cover glass should be placed over the insect,
resting on three or four, dots of thick balsam. In this paper the
descriptions and drawings of the immature stages are based princi
pally on fresh specimens mounted as above.

* Transactions New Zealand Institute, vol. xxviii, p. 415.

IO

For permanent mounts two methods may be used with success.
The first is especially desirable for pupae and pupa cases, particu-
larly when these are provided with long dorsal wax rods, as in the
two species treated in this paper. A small dot of balsam is placed
in the center of a ring of very thick balsam, which should be three-
quarters of an inch in diameter and varying in height according to
the height of the specimen and the length of the dorsal wax rods.
A pupa or pupa case is then placed on the dot of balsam in the
center of the ring. Some annoyance may be saved by examining
the specimen at this time to make sure that it is uninjured. If a
living pupa is used, the slide should now be placed in a strong cyan-
ide bottle for at least fifteen minutes to insure against the mount
being spoiled by the emergence of the adult. A circular three-
quarters inch cover glass should be placed on the ring of balsam and
sealed, so as to make the chamber air tight. I have found chloro-
form balsam more satisfactory than xylol balsam for the ring, as the
former hardens much more quickly. If the balsam is not thick
enough to prevent the cover glass from settling, three cubical blocks
of wood of suitable size may' be imbedded in it. If preferred, zinc
cement may be used in the place of the thick balsam. Specimens
of the younger stages may be mounted in a similar manner, except
that they should not be fastened to the slide, and before sealing the
cover glass they should be gently heated over a flame, until all the
moisture is driven off and nothing remains but the dried bodies.
The results are very uncertain, however, for if the heat is applied too
rapidly the bodies will shrivel.

The writer has found the following simple method the best for
preparing permanent mounts of the first, second and third instars.
Put a drop of xylol on the middle of a clean slide, remove several
living specimens — of all three larval instars if desired — directly from
a leaf and place them in the xylol. From time to time for from five
to ten minutes, xylol should be added to make up for the loss by
evaporation, and when the insects are apparently well cleared the
xylol can be drawn off from one side by a piece of filter paper, leav-
ing the specimens in the middle of the slide. A drop of thin xylol
balsam should now be placed immediately on the specimens, and a
cover glass added. If by examination with a compound microscope
the sides of the larvae are found to be curled inwardly, a slight pres-

1 1

sure on the cover glass will flatten them out to their natural form.

Pupae may be mounted in xylol balsam by dehydrating them in
xylol for a few hours or a day.

With adults the best results are obtained by a somewhat longer
process of dehydration. They may be killed in hot water or in a
cyanide bottle. They are then placed in thirty-five per cent alcohol,
a few drops of ether on the surface of which will dissolve the wax
from the wings and bodies of the insects and allow them to sink.
After an hour the specimens are run successively through the follow-
ing : Fifty, seventy, eighty, ninety, ninety-five and absolute alcohol ;
then an equal mixture of absolute alcohol and xylol ; and finally,
pure xylol, leaving them at least an hour in each.

The alcohol and xylol should be kept in small vials, tightly
stoppered and correctly labelled. The specimens may be removed
from one vial to another by means of a pipette or dropper and a
toothpick, the latter acting as intermediary ; the specimens being de-
posited on the end of the toothpick and then placed in the next vial
in the series. This allows the transference of the specimens with a
minimum amount of the lower percentage alcohol being transferred
at the same time to the higher. An equally satisfactory, and per-
haps less troublesome, method is to keep the specimens in the same
vial and to replace the dehydrating media, successively returning
each in due time to its stock vial.

For the study of the tracheal system, specimens of the imma-
ture stages may be mounted in xylol by the ordinary method, except
that they should not be left more than five minutes dehydrating in
xylol. If a large number of specimens are mounted, some of them
will almost invariably show the main branches of the tracheal system
as conspicuous dark lines, owing to the presence of air in them.

For the study of the life history, leaves of plants with eggs or
young attached, may be kept for two or three weeks in tightly stop-
pered bottles containing a little wet cotton. If moisture collects on
the sides of the bottle, the cork should be removed for a short time
each day, otherwise the leaves will soon turn dark and decay.

In determining the results of treatment with insecticides, posi-
tive conclusions within a few days after the treatment is applied, as
to whether or not the young insects are alive, are difficult to obtain.
When placed upon their backs, the live larvae and pupae usually

12

show some movement of the mouth setae. When examined from the
dorsal side with transmitted light, living specimens usually show more
or less movement of their internal organs. More satisfactory exami-
nation can be made with a hand lens several days after the treatment,
when the dead insects are brownish in color.

The Greenhouse Aleyrodes, Aleyrodes vapora-

riorum Westw.

DESCRIPTION.

Egg. (Plate II, Figs, i and 2.)

The egg is irregularly ovoid, the apical end being the more
pointed, and one side more or less flattened. On the basal end,
usually a little to one side of' the center, toward the more rounded
side, is a short, slender stalk, which is inserted in the leaf. The
length of the stalk varies from one-fourth to one-eighth the length of
the egg, its diameter being about .01 mm. Its form varies, being
seemingly determined by the epidermal cells of the leaf between
which it is inserted, and it is frequently encircled near its base by a
transverse ridge. Color of egg when first laid, light yellowish green,
glistening ; chorion very soft and delicate, somewhat viscid. The
eggs are translucent at first, but gradually become darker, until
finally, after two or three days, they are opaque. At this time the
chorion has become quite hard, and it is difficult to detach an egg
from the leaf without injuring it. In the course of a few days the
egg, by contact with the adults, usually becomes more or less covered
with a flour-like substance. Parts of the egg not so covered appear
of a metallic bronze color. The surface of the egg is unmarked, but
within and usually about in the center, may be seen a rounded,
orange colored body, surrounded by colorless yolk globules. The
developing embryo can be plainly seen.

The length of the eggs, exclusive of the stalk, varies from .187
to .236 mm. ; the greatest transverse diameter varies from .077 to

*3

.11 mm. Average'measurements of twenty eggs from several differ-
ent females gave : Length, .209 mm. ; greatest transverse diameter,
.0901 mm.

First Instar. (Plate II, Figs. 3, 4 and 5. Plate III, Fig. 9.)

In the first instar, the insect is oval in general form, the anterior
end being the more broadly rounded, the sides of the thoracic region
being approximately parallel and the abdomen narrowing posteriorly.
A longitudinal rounded dorsal ridge about one-third as wide as the
body, runs nearly the entire length. On each side of this ridge is a
flattened area considerably thicker than the narrow, thin rim which
surrounds the whole. The change in thickness of the body at the
inner edge of the thin marginal rim is quite abrupt. The dorsal
surface of the body also shows irregular elevations, depressions and
sutures, the latter more or less distinct, marking the thoracic and ab-
dominal divisions. Except in specimens mounted in balsam, these
divisions do not appear to cross the thin marginal rim. The body
as a whole is quite thin at first, but before the first moult takes
place it becomes well rounded above, the dorsal ridge disappearing
at the same time. From the under side of the thin marginal rim
many minute glistening granules may be seen, which appear to be
arranged in rows. The margin of the body is entire, except for the
attachment of the spines. Eighteen spines arise on or near the mar-
gin of the body on each side, and all except numbers two and sixteen,
counting from the anterior end, arise from the extreme margin.
Number two arises on the under side of the thin marginal rim a little
behind number one. Number sixteen arises just inside the margin
on the ventral side, a little nearer number seventeen than fifteen.
The distances between the base of numbers three and four and four
and five are about equal, and greater than between any other adja-
cent spines. Except numbers two, four and sixteen, the spines are
situated at nearly regular intervals. Numbers one, three and four
are horizontal and curved anteriorly ; number two is directed down-
ward and curves inwardly. The remaining spines are horizontal,
and excepting numbers sixteen and eighteen curve posteriorly. Num-
ber sixteen curves outward and downward, and number eighteen, in
respect to its mate, diverges at the base and converges outwardly,
the extreme tip being extremely slender and inconstant in position.

14

Corresponding spines in different individuals vary slightly in length,
but the following figures indicate about the average proportionate
lengths. To obtain true lengths in millimeters multiply the figures
given by .0035.

I

2

3

4

5

6

7

8

9

10

1 1

12

J3

14

7

"8"'

8

' ~6~'

H'

>
5

>
4

5

0

X'

H'

~¥

3'

>

4

41

15

16

i7

18

5

1 +

7

44*

A marginal wax secretion appears soon after the young larva
settles down. This arises from minute marginal pores, in the form of
narrow ribbons, which unite laterally and form a continuous fringe,
the outer edge of which is ragged. This wax is dull, translucent,
and contains many opaque brownish granules. The width of the
fringe is never more than one-fourth the width of the body.

The divisions of the thorax are not clearly marked, and it is a
question where the head ends and the thorax begins. The abdomen
shows eight and possibly nine segments, the last two or three being
greatly modified on account of the vasiform orifice.

This (the vasiform orifice) is about as wide as long, its form
being similar to an equilateral triangle with rounded corners (Plate
II, Fig. 5.) The entire length from front to rear is about one-tenth
the length of the body. The operculum is sub-elliptical in outline,
flattened on the basal side. The posterior margin under high power
objectives, shows two tooth-like projections — one on each side — be-
tween which minute spines can be seen, which appear to extend
around on to the under surface of the operculum. The lingula
(=lingua) is spatulate in outline, bearing eight longitudinal rows of
minute setae above, and two pairs of spines on the caudo-lateral
margin, which curve upward and backward. The posterior ones are
slightly longer than the others, and are about one-fifth as long as the
orifice. The orifice is bounded laterally by chitinous thickenings
which do not unite posteriorly. The posterior end of the orifice
reaches in this instar nearly to the caudal margin of the body.
Just inside the apex of the orifice is a glistening, crescent-shaped
structure, whose convex side is directed posteriorly.

There are two pairs of simple, reddish-brown eyes — a dorsal
and a ventral pair — situated nearly opposite each other just mesad

i5

to the thin marginal rim and about equidistant from the fourth and
fifth spines on their respective sides of the body. These eyes are
rounded and less than .01 mm. in diameter.

There are three pairs of dorsal spines. The first pair is situated
on the cephalic region, one on each side, a short distance mesad
from the eyes. The second pair is situated one on each side of the
basal abdominal region, apparently on the third abdominal segment.
These two pairs of spines are very minute, and require a high power
objective for their detection. The third pair of dorsal spines is
much larger, and situated one on each side of and a little anterior to
the operculum of the vasiform orifice.

On the ventral side of the body (Plate II, Fig. 4) are the legs,
antennae and mouth parts. The entire length of a leg (Plate III,
Fig. 9) when straightened is about one-half the width of the body.
The coxa? are short and stout, and near the base of each of the two
posterior pairs on the inner side is a spine about as long as the
diameter of the coxa. These spines are usually directed inward and
backward. The trochanters are short ; those of the anterior pair of
legs are sub-cylindrical, about one-third as long as wide ; those of
the two posterior pairs of legs appear more or less hoof-shaped, and
each of the six trochanters bears a short spine anteriorly. The femur
is about twice as long as the coxa and trochanter together, sub-
cylindrical in form, tapering toward its outer end. The tibia is a
little longer than the femur and more slender, in each of the two pos-
terior pairs of legs bearing on its outer side near its base a spine as
long as the whole tibia itself. This extends obliquely outward, and
is usually curved near its tip. In addition, all three pairs of tibise
bear a number of minute spines. The tarsus, which consists of a
single segment, is short, knobbed at the tip, with a stout curved
spine about half as long as that borne on each of the two posterior
tibiae, arising on its outer side near its base.

Half way between the first and second pairs of legs in the mid-
dle line of the body is a conical fleshy papilla — the rostrum — from
an opening in the apex of which, surrounded by four minute spines,
the mouth setae protrude. The length of the setae varies, but when
bent backward they usually extend to a point between the hind
coxae and the caudal margin of the body. In front of the rostrum
is a prostomial plate or shield, subovate in form, the broader end

i6

being anterior. It is truncate where it touches the base of the mouth
papillae, slightly concave on the sides posteriorly, broadly rounded
anteriorly with two movable papillae on the anterior margin, each of
which bears a long spine, about equal in length to those borne on
the coxae of the two posterior pairs of legs. A curving suture on
each side of the anterior two-thirds of the plate divides it into three
parts, a long central piece with two side pieces. On the ventral sur-
face of the abdomen, underneath the operculum, is a pair of spines,
one on each side, about equal in length to those which arise at the
anterior end of the prostomial plate. These spines extend back-
ward, usually reaching nearly to the caudal margin of the body.
Anterior to and outside of the base of each of these spines there is
a small pore, which probably represents the opening of some gland,
as an amorphous, waxy (?) substance usually covers one or both
openings after the larva is two or three days old.

Each antenna arises on a line about half way between the fore
coxae and the anterior margin of the body, and the length is between
one-half and two-thirds the width of the body. Each consists of four
segments, the basal one being short and stout ; the second twice as
long as the first, and more slender, reaching nearly to the margin of
the body between the third and fourth spines of each side, (the an-
tennae usually being directed in that direction) ; the third segment
very short, sub-globose, and bearing two or three short, stout spines ;
the fourth twice as long as the second, slender, its apical third bent
anteriorly, minutely spined, and with a larger and more conspicuous
spine arising at about two-thirds the distance toward the tip on the
posterior side, and another still larger one at the tip. Each ventral
eye is situated outside of and slightly behind the base of the an-
tenna of its respective side of the body.

Segmentation from below is less distinct than from above. The
color of the larva is pale green, semi-transparent, with two internal
orange yellow bodies of irregular rounded form, situated one on each
side in the basal abdominal • region. These sometimes enclose a
clear area. The tracheal openings are ventral and very difficult to
distinguish in this instar, but their positions are marked by the gran-
ular appearance of the surrounding regions and by the abrupt termi-
nations of the main branches of the tracheal system, which maybe
more or less plainly seen in the semi-transparent body of a newly

i7

hatched larva. Spiracle number one is located between and outside
of the base of the second and third pairs of legs ; number two be-
tween the base of the second and third pairs of legs; number three
a little behind the base of the third pair of legs, and number four a
little behind the ventral spines in the caudal region.

The length of the body in this instar varies from .26 to .30 mm.;
average length about .28 mm.; the greatest width from .121 to
.165 mm.; average greatest width about .16 mm.

Second Instar. (Plate II, Figs. 6, 7 and 8; Plate III, Fig. 10.)

In this stage the outline is more variable than in the first, vary-
ing from broadly oval to elliptical, usually with a slight inward curve
on each side of the thoracic region. The margin is finely crenulate,
but there is no well marked marginal rim as in the first instar. Im-
mediately after moulting, the body is flat and thin, but before the
next moult it becomes well rounded above. Three pairs of marginal
spines are present. The first pair is on the cephalic margin, one on
each side ; the second pair on the caudo-lateral region, one on each
side, and the third pair on the caudal margin. These probably rep-
resent spines numbers one, sixteen and eighteen respectively, of the
first instar. The third pair is a little more than one-tenth the length
of the body, the second pair is a little more than one-third the length
of the third pair, and the first pair is very minute, sometimes appar-
ently lacking. The first and second pairs usually become obscured
soon after the ecdysis by the lateral wax fringe, which, however,
never extends much beyond the basal half of the third pair. There
are three pairs of dorsal spines ; the first pair is on the cephalic
region, as in the previous instar ; the second pair is on the first ab-
dominal segment, one on each side ; and the third pair is near the
vasiform orifice, one on each side, opposite the operculum. Of these
spines, the second pair is invariably minute, as in the first instar,
but the first and third pairs, while very variable in different indi-
viduals, are developed to a much greater degree than in the preced-
ing instar, being sometimes as much as one-fourth longer than the
spines on the caudal margin, or they may be but one-fourth as long
as these spines. The variation in length of the two pairs of spines
is apparently independent ; in one specimen the first may be more
or less longer than the third pair, while in another the third may be

iS

somewhat longer than the first pair, and, again, both pairs on the
same individual may approach the maximum or the minimum in
length. Toward the end of this stage, traces of the first and third
pairs of dorsal spines of the third instar may be observed beneath
the dorsal surface of the body. The first pair of spines arise directly
beneath the bases of their ontogenetic predecessors, and are directed
forward obliquely so that the two cross near their tips, in the middle
line of the body. The developing third pair of spines are directed
backward and outward, and like the first pair arise beneath the base
of their predecessors. The segmentation of the abdomen is fairly
distinct in the middle, while that of the thorax is more obscure.

The vasiform orifice (Plate III, Fig. 10) is relatively farther for-
ward in this instar than in the preceding one, which is indicated by
the comparatively greater distance from the apex of the orifice to the
caudal margin of the body, and by the fact that the spines on the
dorsum near the orifice now lie opposite the operculum, instead of
anterior to it, as in the first instar. The vasiform orifice is of about
the same general form as in the first instar. The operculum lacks
the tooth-like process on each side of the caudal margin, and is pro-
portionately longer than before, extending nearly one-half the dis-
tance toward the apex of the orifice. The lingula is spatulate, with
two pairs of side lobes and one terminal lobe. When in its natural
position its caudal end reaches to about three-fourths the distance
from the anterior to the posterior end of the orifice. On each side
of the terminal lobe arises a spine which extends backward to a little
beyond the apex of the orifice. A smaller spine arises on each side
between the two side lobes. The upper surface of the lingula bears
about fourteen longitudinal rows of minute seta;. The apical half
of the operculum and the lingula are more opaque than the other
parts of the body. The chitinous ridges which bound the orifice
laterally do not meet behind, although the intervening space between
their ends is comparatively smaller than in the previous instar.

The eyes are proportionally smaller than before, and are situated
internally instead of at the surface as in the first instar. The eyes
on each side are about on a line with and outside of the two dorsal
spines (first pair) on the cephalic region.

On the ventral side of the body (Plate II, Fig. 7) the vestigial
legs and antenna; can be plainly seen, their relative position being

J9

as before. The antenna; (Plate II, Fig. 8) are directed backward,
and extend a little more than half way to the base of the fore legs.
They are rather thick at the base, gradually tapering to the apex and
bearing numerous blunt spines or papillae. Of these, two near the
base, one on each side, are especially prominent in most specimens.
Two segments can usually be distinguished in the antennae, the
basal segment being about one-third as long as the apical one. Oc-
casionally the basal third of the apical segment appears to be cut off
as a third segment. The antenna? are immovable, or practically so,
in this stage as well as the following immature stages. The legs are
a little longer than the antennas and of the form of a truncated cone,
transversely wrinkled, with no distinct segments, and terminate in a
rounded knob, which probably has an adhesive function. As a
whole, the legs are suggestive of the prolegs of caterpillars. A trans-
verse wrinkle which usually appears at about two-thirds the distance
from the base to the tip, possibly indicates the line of division be-
tween two segments. A pair of minute spines is usually found on
the basal parts of the second and third pairs of legs, one on the inner
and one on the outer side. The external mouth organs appear as in
the first instar. The anterior end of the prostomial plate is indis-
tinct, and the pair of spines which occurs there in the first instar is
now wanting. The pair of spines on the ventral surface below the
operculum is present as before. Near these in some specimens the
oval gland openings can be seen, their relative position being as
before. Each spiracle, or tracheal opening, of the two anterior pairs
appears to be double in this instar, consisting of a moderately large
opening with a smaller one directly behind it.

The color of the body is the same as in the first instar ; the
length varies from .341 to .407 mm. ; the width varies from .189 to
.235 mm.

Third Instar. (Plate III, Figs. 11 and 12.J

In this instar the form, marginal and dorsal spines, marginal
wax secretion, ventral spines and color of the body are as in the sec-
ond instar. The legs are of about the same proportionate length,
while the eyes are relatively smaller than before. The vasiform
orifice is longer than wide, resembling a triangle with rounded cor-
ners in form, its apex distinctly indented. The lateral chitinous

20

ridges do not unite posteriorly, but are connected by a median,
crescent-shaped thickening. The operculum is nearly semi-circular,
reaching to about one-half the distance from the base to the apex of
the orifice, and the lingula has its side and terminal lobes more dis-
tinctly marked than before. The number of longitudinal rows of
minute setae on the upper surface of the lingula is now about
eighteen. A shallow groove extends from the apex of the orifice to
the caudal margin of the body. The antennae arise nearer the base
of the fore legs than in the previous instars, and may be partly con-
cealed by them. They are indistinctly segmented, thick at the base,
tapering toward the tip. The basal two-thirds of each is directed
inward toward the antenna of the opposite side, while the apical
third is bent backward toward the base, the whole forming a figure
not unlike the letter C. The first two pairs of spiracles now appear
as single slits, while the other two pairs are indistinct.

The length of the body in this instar varies from .493 to .583
mm., the width from .275 to .352 mm.

Pupa. (Plate III, Figs. 13 and 14; Plate IV, Figs. 15 and 16;

Plate VI, Figs. 2S-31.

The form of the pupa varies from irregularly oval to elliptical,
the broadest part- of the body being usually about two-thirds of the
distance from the cephalic to the caudal margin. When freshly
moulted the body is flat and thin, with no wax secretion, but as the
insect grows it becomes raised from the surface of the leaf by a ver-
tical wax fringe, the height of the body becoming about one-third of
its width. The dorsum is rugose, less noticeably so along the mid-
dle ; somewhat convex. There are three pairs of marginal spines as
in the two preceding instars. The first pair on the cephalic margin
are minute, as in the two previous instars, and frequently appear to
be lacking ; the caudal pair, arising a little inside the margin, are
less than one-tenth the length of the body, and the spines on the
caudo-lateral margin are less than one-half as long as the last men-
tioned. The caudal pair curve upward and backward, diverging at
the base, usually converging posteriorly. There are three pairs of
dorsal spines, as in previous instars. The first two pairs are minute
— the first pair being somewhat longer than the second. The third
pair is usually well developed, varying greatly in length in different

21

individuals, though always longer than the other dorsal spines. In
one specimen examined these were observed to be about one-third
longer than the vasiform orifice, representing about the maximum
length.

Eight abdominal segments are evident from above, except on
the extreme sides ; that which appears to represent the ninth seg-
ment is much modified by the vasiform orifice being much narrowed
posteriorly and somewhat flask-shaped in outline. This portion of
the dorsum is much more clearly defined than in previous instars.
The vasiform orifice (Plate III, Fig. 13) is similar in form to that of
the previous instar. The chitinous ridges which bound it laterally
appear to meet posteriorly, with a rounded chitinous thickening at
the point of union. The operculum is nearly hemi-elliptical in out-
line (having the form of an ellipse cut through its shortest axis),
reaching from the base of the orifice to a little over one-half of the
distance toward the apex. The lingula has one large apical lobe
and three pairs of smaller side lobes, and is densely covered with
longitudinal rows (about twenty-five) of minute setae. From each
side of the apical lobe below arises a spine, which curves slightly up-
ward and extends caudad beyond the apex of the orifice, its length
being somewhat variable in different individuals, but never more than
one-third the greatest width of the vasiform orifice. A second pair,
less than one-fifth as long as these, arise, one on each side, between
the first and second side lobes. In one instance a side view of a
pupa case was obtained with the operculum and lingula extending
almost perpendicularly to the dorsum of the case, and the ventral
surface of the operculum and the ventral surface of the lingula
showed minute setae about the size of those present on the dorsal
surface of the lingula. The anterior pair of side lobes of the lingula
is frequently hidden by the operculum. A shallow but well marked
furrow extends caudad from the apex of the orifice to the margin of
the body, where a short cottony tuft of wax almost invariably occurs
in full grown pupae.

The vertical wax fringe is probably homologous to the lateral
fringe of the previous instars, and consists of narrow ribbons of wax
fused together laterally, which arise from the so-called marginal wax
tubes. Above the points determined by C. W. Woodworth* as the

* Canadian Entomologist, vol. xxxiii, p. 173.

3

22

outer ends of the breathing folds in Aleyrodes citri (viz.. one on each
side of the body opposite the first pair of spiracles and a single one
at the anal end) the wax tubes are noticeably smaller than usual, and
the ribbons secreted there are narrower than elsewhere. The dorsal
wax secretion consists of a double submarginal series of glassy waxen
rods and a more dorsal series of from five to eighteen (typically
eight) rods. There seems to be no distinction between these rods
except in size and position of origin. The outer series consists of
from about fifty to seventy-five (fifty-seven to seventy-two being the
limits actually observed) rods, fairly constant in size, in full grown
pupae being almost invariably less than one-fifth the width of the
body in length. They are rather more blunt at the tips than the
other rods, proportionally narrower, and usually curve downward
over the margin. The series which I designate the inner sub-
marginal series is distinguished from the outer series by the much
larger size of the rods. There are typically ten or twelve (five or
six pairs) in this series, but may be as few as five. They usually
arise a little mesad to the rods of the outer series, and with the ex-
ception of the most anterior pajr appear never to arise farther away
from the outer series than the width of their bases. The anterior
pair is almost invariably within twice the diameter of their base from
the outer series. Typically, the inner series of rods consists of three
on each side in front of the mesothorax ; one on each side near the
base of the abdomen ; one on each side nearly opposite the vasiform
orifice, and one on each side of the groove mentioned above which
connects the apex of the vasiform orifice with the caudal margin of
the body. In full grown pupae the rods of the inner series are from
five to ten times as long as those of the outer series. They are
usually directed upward and curved inward over the dorsum. The
rods of the dorsal series, typically eight incumber, are of about the
same size as those of the inner sub-marginal series, and arise as fol-
lows : The first pair on the cephalic region, one on each side about
one-fourth the distance from the cephalic margin of the body to the
base of the abdomen ; the second pair, one on each side, about half
way from the cephalic margin of the body to the base of the abdo-
men ; the third and fourth pairs on the sides of the third and fourth
abdominal segments respectively. The dorsal rods are subject to a
great deal of variation (Plate VI, Figs. 28-30), the most frequent be-

23

ing the addition of rods on the fifth and sixth abdominal segments
corresponding to those on the sides of the third and fourth. The
most remarkable variation seen by the writer was a small rod pro-
duced on the middle line of the body just anterior to the vasiform
orifice. All the wax rods arise from conical bases, and appear to be
formed in sections, each section being sub-cylindrical, tapering grad-
ually outward and ending in a cone, at the inner end of which is a
conical cavity which fits over the outer end of the section directly
behind it and nearer the surface of the body. Through the middle
of each wax rod runs a narrow cavity. This structure is plainly
seen, both directly under a microscope and by dropping a few of the
wax rods into a drop of xylol on a glass slide and observing them as
they dissolve.

A little outside of the outer submarginal series of rods, and
alternating with them, is a row of minute pores, which vary in num-
ber, extending all around except at the caudal end. Sometimes one
is absent and sometimes two occur where but one is normally pres-
ent. Farther up on the dorsum there are about fifteen or sixteen
pairs of similar pores. In general, it may be said that there is a
double row of these on each side of the middle of the dorsum. None
have been found by the writer on the second or the ninth abdominal
segments. Other pores than those included in the above groups may
sometimes be seen. Usually near each pore is a glistening point,
which may be a minute seta, as it appears very much like the struc-
tures found in Aleyrodes fernaldi,* which are, however, many times
larger.

On the venter (Plate III, Fig. 14) the legs, antennae and mouth
parts are distinguishable with some difficulty, except in specimens
that have recently moulted. The legs and mouth parts are as in the
previous in star. The antennas now lie partly hidden in pockets sit-
uated, one on each side, just outside of the anterior pair of legs.
They are directed backward and are straight, quite thick at the base,
gradually tapering toward the tip, where they are abruptly narrowed
and end in a short pointed process. Transverse wrinkles may be
seen, but there is no distinct segmentation. A pair of spines occurs
on the ventral surface, one on each side, below the operculum, as in
previous instars. No eyes can be distinguished in freshly moulted

* Psyche, vol. x, p. 84.

24

specimens, but as the pupa matures the imaginal eyes appear as two
reddish spots in the cephalic region.

The pups are greenish white in color, wiith yellow bodies pres-
ent in the basal abdominal region as in the previous instars ; the
empty pupa cases are white.

The length of the pupa varies from .66 to .88 mm.; the greatest
width from .396 to .55 mm.; average length about .76 mm.; average
width about .49 mm.

Adult. (Plate IV, Figs. 17-20; Plate V, Figs. 21-27.)

Female. The length of the body of the adult female varies from 1 .
to 1.3 mm., the average being about 1.14 mm. The color of the head
and thorax is pale yellowish buff, and of the abdomen pale lemon
yellow. The tip of the rostrum* is black ; the legs and antennas are
pale yellowish, sometimes slightly tinged with dark white. The vasi-
form orifice and the base of the ovipositor may be more or less dark
colored. In specimens mounted in balsam the thorax is deep orange
in color, due to muscles within ; the abdomen is bright lemon yellow,
showing bright orange or orange red ovaries and egg nuclei. Soon
after emergence from the pupa case the whole body becomes covered
with a white amorphous waxy secretion. This first appears in the
form of very fine threads exuded from extremely numerous and
minute pores in the integument.

The head is transverse ; seen from the front, sub-triangular,
rounded above. Two pairs of reddish eyes are borne on each extreme
side, the upper pair of which is wine colored, being a little lighter in
color than the lower pair. There are about forty-eight pigmented
facets to each of the upper eyes, surrounded by about ten unpig-
mented ones. The lower eyes are each composed of about thirty
pigmented facets, larger in size than the facets of the upper eyes.
Both the upper and lower eyes are sub-circular in outline, separated
by a narrow strip of integument. Directly above each upper eye is
a single unpigmented ocellus. Each antenna arises a short distance
in front of the upper eye of its respective side. It consists of seven

* I have here used " rostrum " in the sense of beak. Some writers follow Maskell in
designating this organ the mentum. As Maskell gives no evidence in support of this view,
and as I have been unable to find any reference to a study of the homology of the mouth
parts of this group of insects, I prefer to retain for the present the word " rostrum," as used
by the older writers.

25

segments (Plate V, Fig. 24.) ; the first sub-rotund ; the second club
shaped ; the remainder slender. The following is the usual formula
of the comparative length of the segments :*
2 — 5 — 10 1-2 — 4 — 6 — 4 — 4.

The following formula shows the range of variation observed,
omitting the two basal segments, which are fairly constant and which
can seldom be measured accurately in mounted specimens :

(9-1 1)— 4— (5-6)— (4-5)— (4-5).

The entire length of the antenna is about one-third the length
of the body. The second segment bears a few short spines, and the
apical segment bears a short spine at its tip. The third, fourth,
fifth and seventh segments have one or more sense organs near their
outer ends. These seem to be always absent on the sixth segment.
The third to seventh segments, inclusive, are annulated.

Immediately below the eyes is a short sub-conical structure
(rostrum of Maskell), from the apex of which the mouth setae appear
to arise. The rostrum (mentum of Maskell) consists of four free
segments and a basal apparently immovable portion, which is at-
tached to the body seemingly to the prothorax between the bases of
the fore legs. The basal portion reaches about to the tip of the coni-
cal structure (rostrum of Maskell), and the four following segments
show the following proportionate lengths : 4 — 7 — 5 — 12. The ros-
trum bears a few scattering hairs, which are more numerous on the
terminal segment. The labrum, which is very slender at the tip,
reaches to the outer end of the first movable segment of the ros-
trum.

The thorax is rather short, compact and well rounded above.
The prothorax is slightly smaller than the metathorax, which in turn
is smaller than the mesothorax. The sclerites which compose these
divisions of the thorax are well fused. The two anterior pairs of
legs are of about equal length. The fore coxae are slightly longer
than the middle coxae ; the trochanters are short ; the femora about
two-thirds as long as the tibiae and about equal to the two tarsal seg-
ments together ; the first tarsal segment is about one-fourth longer
than the second. The hind coxae are very stout, with a prominent
infolding anteriorly. The hind trochanters are short, each bearing
a single spine at the bottom of a groove-like cavity on the caudal

* Measurements made with i in. eyepiece and 1-2 in. objective; tube length 145 mm.

26

side. The remaining segments of these legs are longer than in the
two anterior pairs, but their relative lengths are about the same.
All three pairs of legs, more particularly the last two pairs, bear
numerous spines on the femora, tibiae and tarsi. These spines are
rather scattering, with the exception of a row of about fifteen on the
inner side of the hind tibia, an oblique row of four or five spines on
the outer sides of the middle and hind tibiae, a little beyond the
middle of the segment, and about a half dozen spines near the tip
of the tibiae of all^three.: pairs of legs. From the upper side of the
tip of the last tarsal segment (Plate IV, Fig. 17) a long slender spine
arises from a common tubercle. The two curved tarsal claws ap-
parently unite at their bases, and articulate on the lower side of the
tip of this segment ; between the two claws there is a short knife-
blade-like process, pointed at the tip and bearing on the under side
a fringe of very delicate hairs.

The fore wings arise far back on the sides of the mesonotum.
They are about as long as the entire length of the body when the ab-
domen is fully extended, or about 1.15 mm. long and .44 mm. broad
at the widest part. The hind wings arise from the sides of the
metanotum anteriorly, and are smaller than the anterior pair. Both
wings are provided with a single median vein, that of the posterior
wing being nearly straight ; that of the anterior wing being bent
toward the posterior margin of the wing at a point slightly beyond
the middle. A fold in the fore wing, appearing like a branch of the
median vein, is frequently seen arising near its base and extending
obliquely toward the posterior margin. The wings are beaded on
the margin (Plate IV, Fig. 18), each bead consisting of a minute
globule, from the outer side of which two or three minute setae arise.
Three or four slender spines arise near the base of the hind wing on
the costal margin.

The abdomen (Plate IV, Fig. 20; Plate V, Figs. 21-23) *s
more or less spindle-shaped, and consists of eight segments. The
basal segment is small and transverse, sharply separated from the
remainder of the abdomen, in which the segmentation is difficult to
distinguish except in freshly emerged specimens. The dorsal and
ventral plates of the second to the seventh segments inclusive are
separated by a pleural membrane, which is capable of considerable
extension. In specimens mounted in balsam the ventral plates of

27

the third and fourth segments are clearly outlined, the remaining
plates being indistinct. The eighth segment is quite large, and is
terminated by a short conical ovipositor. This organ (Plate V,
Fig. 27) consists of three pieces, and is surrounded near its base by
about eight tactile hairs. Near the base of the eighth segment
above, is the vasiform orifice, which is sub-circular in outline. The
operculum is sub-quadrate, the caudal margin being concave. The
lingula protrudes caudad beyond the orifice ; is strap-shaped, nar-
rowing toward its base, and is minutely setose. The ovipositor is
usually bent upward when not in use, and sometimes appears to be
held in a vertical position by a projecting fold of the eighth seg-
meut. A pair of minute spines is borne on the under side of the
second abdominal segment near its base ; a minute pair is present
on the dorsum of the fifth, sixth and seventh segments ; two pairs
on the dorsum of the eighth segment, and a few spines on the sides
a little anterior to the base of the ovipositor. These minute spines
are all, as a rule, indistinguishable in specimens mounted in balsam.
The writer has not been able to positively locate the spiracles.

Male. The male (Plate IV, Fig. 19 ; Plate V, Figs. 25 and 26)
differs from the female only in size, and in the form of the genitalia
and of the abdomen. The length varies from .9 to 1.1 mm., the
average length being about .95 mm. The number of abdominal
segments is nine, the vasiform orifice and the genitalia being upon
the ninth instead of the eighth segment as in the female. The third
to sixth ventral plates of the abdomen, inclusive, are quite promi-
nent in outline, especially in specimens mounted in balsam. Geni-
talia (Plate V, Figs. 25 and 26) forcipate, consisting of two side
pieces or claspers and a median penis, the former being provided
with tactile hairs. Either the right or left clasper, differing in dif-
ferent individuals, appears to have a dorsal swelling near its base,
which shows only in a side view. From the side the penis is large
at the base, gradually tapering toward the tip. There is a marked
upward curve near the base and another less marked upward curve
near the tip. From above, the penis appears barrel-shaped at its
basal fourth, outwardly becoming flattened laterally ; the outer half
or two-thirds is thin, the sides being nearly parallel. The opening
at the tip of the penis is circular. It is usually bent upward, and
consequently is not seen in its full length from above, but a side

28

view (Plate V, Fig. 28) shows that it is about as long as the
claspers.

Tracheal System of the Immature Insect.

The diagram on Plate VI (Fig. 32) illustrates the main branches
of the tracheal system of larval and pupal Aleyrodes, based on a
study of the two species, vaporariorum and packardi. Prof. C. W.
Woodworth has worked out more in detail the tracheal system of A
citri.* It will be seen that the diagram here given differs from his des-
cription only in a few minor details. What he has called " dorsal gir-
dles " I find to be ventral, and his " ventral trunks " I find to be
dorsal in the species examined. I have not found the branches
which he describes as arising from the transverse connections be-
tween the longitudinal trunks and the third pair of spiracles. The
following is the explanation of the letters used in the diagram :

I II III
V V V

Ventral loops.

D D D

Dorsal trunk.

I II III IV

T T T T

Position of spiracles.

L L L

Vestigial legs.

R

Rostrum.

E E

Eyes.

A A

Antennae.

NOTES ON LIFE HISTORY AND HABITS.

Egg. The mechanical condition of the leaf seems to have some
influence on the vitality of the eggs, for if a leaf upon which the
eggs have been deposited within five or six days is allowed to wither
and become dry, the eggs will not hatch. Those eggs which are
nearly mature are not so affected. As a rule, the eggs hatch in from
ten to twelve days, though this period may be prolonged by low
temperature. On hatching, the egg splits longitudinally from apex
to base and the larva slowly frees itself. The egg shell collapses at
the same time that the larva escapes, the free edges of the shell
curving inward.

Larva. The newly hatched larva is quite active and may

* Can. Ent., vol. xxxiii, p. 173.

29

crawl some distance before settling down, or may remain quite near
its place of birth. Its rate of locomotion on a leaf is between one
and two millimeters per minute, equivalent to about one-half an inch
in ten minutes. The larvae rarely crawl in a straight line, however, as
they double on their own tracks and change direction at every little ob-
struction such as a hair or a particle of dirt on a leaf, so that it is a
rare accident if one finally reaches a distance of one-half inch before
settling down, where it remains until it reaches the adult condition.
Everything considered, there seems to be little chance of an Aley-
rodes in this instar getting from plant to another, or even to another
leaf, unless the leaves are in actaal contact. The young larva usu-
ally settles down within twenty-four hours and gradually loses the
use of its legs.

The duration of the first instar is from five to seven days, of the
second and third instars from four to six clays each, and of the fourth
or pupal instar from thirteen to sixteen days. At moulting, in the
first three instars, the skin splits apparently around the anterior
margin of the body, and is then gradually moved back, aided by up
and down movements of the abdomen, and usually drops off entirely
unless entangled by the hairs of the leaf. Moulting appears to be
a slow process, from two or three hours to a whole day being re-
quired before the change is entirely completed. As each portion of
the body becomes freed from the skin it spreads out over the surface
of the leaf and immediately assumes the form and horizontal dimen-
sions which continue throughout the instar.

If the ventral surface of a freshly moulted insect of the second,
third or pupal instars be examined, it will show a large fleshy pro-
jection arising on each side between the bases of the first and second
legs. These structures are concave at the tip, and are probably ad-
hesive in function, serving to keep the insect attached to the leaf
during the process of ecdysis, as they disappear soon after. A
specimen in the second instar, placed on its back on a glass slide,
was observed to withdraw these almost completely within five min-
utes.

Changes in position of the Aleyrodes in the second, third and
pupal instars are very slight, if any, and these occur only at the time
of moulting, or immediately afterward. In one instance, one of
these insects in the third instar was observed to move in the course

3°

of a few minutes so that its long axis formed an angle of ninety de-
grees to its original position.

Between the moults in all the immature stages — considering
hatching from the egg as a moult — lateral growth of the body is not
appreciable, increase in size seeming to result almost entirely from
growth in thickness.

During the last few days of pupal life the insect does not feed or
perceptibly increase in size, and the developing imaginal characters
can be more or less distinctly seen within. It is at this time, only
that the Aleyrodes is a true pupa in the sense the word " pupa " as
used in other Holometabola, or insects with complete metamorphosis ;
thus what is known as the pup a case is in reality the last larval skin.

In all the immature stages there is exuded from the vasiform
orifice at intervals a colorless liquid similar to the honey dew of
aphids. Sometimes this excretion collects in large viscid globules
immediately below the insect secreting it, apparently suspended
either by the hairs on the leaf or by the dorsal spines of the insect.
The cause of the insects feeding almost exclusively on the under
surface of the leaves is apparent. If they should attack the upper
surface of the leaves in large 'numbers they would cause their own
destruction, for in this case the honey dew, instead of dropping to
the leaves below, would spread out over their bodies and serve as
food for fungi, the mycelia of which readily penetrate the tender
skins of the insects themselves.

Adults. The adults emerge through a T like opening in the
pupa case. The integument splits along the middle of the dorsum
from the anterior end of the body to the base of the abdomen,
where it joins a transverse split which follows the line of segmenta-
tion, reaching nearly to the lateral margin of the case on each side.
A newly emerged adult is devoid of wax secretion, the wings are
folded up like crumpled paper, and the legs are delicate and much
twisted. In a short time secretion of wax appears, the wings unfold
and the legs straighten. The wings are rarely used unless the insect
is disturbed, and then only for short, more or less erratic flights. It
is probable that the wind plays an important role in their distribu-
tion out of doors. In the laboratory single specimens on isolated
plants have been observed to remain for days at a time on a single
leaf.

31

The adults are frequently seen in coitu, but probably many eggs
are unfertilized. Before pairing, there seem to be certain prelimi-
nary movements. The male takes a position alongside of the female,
intermittently flaps its wings and at the same time continually strokes
with its own antennae the nearest antenna of the female. Their
position is well illustrated by Davis (17). Frequently two males, one
on each side, thus court a female. If the suit is successful, the
female slightly raises the tip of the abdomen, at the same time the
male taking a position of about thirty or forty degrees with the body
of the female, with its inner wings arising above those of the female,
bends the tip of its abdomen upward, and clasps the genitalia of the
female from below.

Egg Laying. The female seems to have no choice of position
on a plant for deposition of eggs, but leaves them wherever she may
happen to be feeding. Eggs are not uncommonly found on the upper
surface of the leaves, on the petioles and even on the stems of the
plant, though the great majority are found on the favorite feeding
place of the adults — the under surface of the leaf. The adults prefer
the youngest leaves, and there is a slow but continual migration up-
ward to keep pace with the unfolding of the leaf buds, the majority
of the freshly laid eggs being found, therefore, on the upper leaves
of a plant. The female frequently, as has been observed by pre-
vious writers, uses her rostrum as a pivot, and deposits her eggs in
a more or less complete circle about her. From ten to twenty eggs
are often found in one of these circles, which are about 1.5 mm. in
diameter. Eggs are also deposited singly, this being the case es-
pecially on hairy leaves. The writer has observed as many as twelve
thousand eggs per square inch on the under surface of a Salvia leaf
with more being constantly added. The eggs, however abundant,
never touch one another, it being probably impossible for the female,
on account of the conformation of the ovipositor and the end of the
abdomen to deposit eggs so close to one another that they are
actually in contact.

Observations on the duration of adult life, parthenogenesis, etc.
Adult females have been isolated on plants previously free from
Aleyrodes in any stage, for the purpose of determining the duration
of adult life, the number of eggs laid by each female, whether or not
parthenogenesis occurs, and if so, its character. The females iso-

32

lated for the purpose of these observations were seen to emerge from
their pupa cases, and consequently there was no possibility of their
having been fertilized. The plants upon which these females were
kept were growing in small pots covered with lantern chimneys,
which were closed at the top with cheese cloth. Four trials were
made:

i. April 3, 1902, an unfertilized female began egg laying, and
on April 17 three eggs were observed to have hatched.

2. April 17, 2902, an unfertilized female began egg laying, and
on April 29 several eggs had hatched.

3. Dec. 8, 1902, a female emerged, was isolated on a tomato
plant, and began egg laying Dec. 12 (females usually begin egg lay-
ing on the second or third day after emergence), and continued,
averaging four per day for eleven days. Personal observations were
here discontinued, but E. A. Back, an undergraduate student at the
Entomological Laboratory, noted that the adult died Jan. 1, 1903.
On Jan. 7 I found the plant dead, apparently from cold, and on ex-
amination of the leaves I found that about three-fourths of the eggs
had hatched, and that some of -the larvae were in the second instar at
the time the plant died. Quite a number of eggs were found that
had certainly been laid during my absence, but they were not
counted.

4. March 17, 1903, a female emerged from its pupa case, and
was isolated on tomato and chickweed growing in the same pot.
Egg laying began March 18. Eggs were deposited on the stems and
upper and lower surfaces of the leaves of both plants, making it im-
possible to count from day to day all the eggs that had been laid.
On April 2, forty-nine eggs were counted, and on April 22 eighty more
were known to have been added to this number. There were about
eight days altogether when the female was in such a position on the
plant that no attempt was made to count the eggs for fear of dis-
turbing her. At a very low estimate, twenty- five eggs were laid dur-
ing these days, The offspring of this female began to emerge as
adults on April 22, and the original female was transferred to a
chickweed plant growing in another pot. By an accident I lost on
the same day the positive identity of this insect, but I am quite sure
that she produced the forty-nine eggs which I counted on April 29,
after which observations on this insect were discontinued. So far as

33

observed, all the eggs laid by this female hatched and the young
reached maturity, the adults being males without exception.

To summarize these observations, unfertilized eggs hatch and
the larvae develop into adults of the male sex. Two females were
known to lay forty-four and one hundred and twenty-nine eggs re-
spectively, and in both cases many more were undoubtedly laid.
These same insects lived in the adult condition for twenty-three and
more than thirty-six days respectively.

I have tried several times to isolate a female which had cer-
tainly been impregnated, but was unsuccessful. It is not impossible
to do this, however, and I suspect that when this is done the young-
produced from fertilized eggs will all develop into females, giving us
a condition similar to that which is generally believed to occur in
the honey bees, and known as arrhenotoky.

In regard to the length of adult life, I might further add that in
greenhouses where there are millions of live adults on the plants, it
is difficult to find a single dead specimen on the benches, providing
they have not been killed by artificial means. This is a further indi-
cation that natural deaths among adults are rare, and that the adult
life of each individual may extend over many weeks.

Should it prove true that unfertilized eggs of this insect produce
only males and fertilized eggs only females, then the number of adult
males and females will be in direct proportion to the number of un-
fertilized and fertilized eggs. In Psyche (April, 1903) I gave an
estimate of the proportion of the two sexes of Aleyrodes in nature,
based on actual count of eighty-five specimens of adult Aleyrodes
taken at random, representing four different species. The figures
given were twenty males to sixty-five females. For the purpose of
obtaining a more exact idea of the. proportion of the sexes in the
present series I counted one hundred adults taken at random, and
found twenty-three males to seventy-seven females.

The spreading of the adults from greenhouses probably accounts
for the presence of this species on out of door plants in most cases,
but there is no reason to suppose that it cannot pass the winter out
of doors in the egg state as do other species of the genus in this
region.

34

IDENTITY OF THE INSECT.

In regard to the specific identity of this insect, there seems
little room for doubt that it is the Aleyrodes vaporariorum of West-
wood. Although Westwood's description (i) may not be sufficient
for identification, Signoret's description (2) is quite complete, and
the correctness of the latter's determination has never been chal-
lenged by later European writers. On the contrary, J. W. Douglas
(7), who was later the best authority on this group of insects in En-
rope, speaks of Signoret's work in such a manner as to leave no
doubt that the species described by Westwood and re-described by
Signoret were the same.

Signoret's drawings and description of the pupa differ from
those of the present writer only in minor details. His determination
of the limits of the second and third abdominal segments differs
slightly from mine ; consequently, according to his drawing and de-
scription of the pupa a pair of wax rods arises from the second ab-
dominal segment, whereas I found the corresponding rods on the
third abdominal segment. His drawing of the ventral side of the
first instar is obviously inaccurate in many respects, but it probably
resembles the first instar of Aleyrodes vaporariorum quite as much
as of any other species.

ORIGIN AND DISTRIBUTION.

This insect was first described in 1856 by Westwood (1), who
supposed it to have been imported into England from Mexico.
Signoret, in 1868 (2), intimated that the species might have been in-
troduced into Europe from Brazil. Whatever its origin may have
been, it is at present widely distributed in greenhouses in Europe
and in the Northeastern United States, and has also recently been
reported from Canada (27).

FOOD PLANTS.

The list of food plants of the Greenhouse Aleyrodes is so ex-
tensive that it would seem almost advisable to list only those plants
which the adults reject and upon which the immature stages cannot
subsist. Westwood (1) in his description of the species says in re-
gard to its food plants : " It especially attacks the leaves of Mexi-

35

can species of Gonolobus, Tecomia vclulina, Bignonice, Aplulaiuh a ,
Solamnus and other similar soft-leaved plants." Signoret (2) found
it on Salvia splendens and Tancana camara, and Douglas (7) speaks
of its injuries to cucumbers and tomatoes. Quaintance (19) reports
"what appears to be this insect" on Fuchsia, Pelargonium and
Oxalis from various parts of the Eastern United States. Britton
(23 and 27) lists fifty-eight food plants in Connecticut upon which
the insect was observed in its immature stages. It has recently been
recorded on violet (26).

It is sufficient to say here that the Greenhouse Aleyrodes is a
very generous feeder, its food plants representing several families
and orders. The following food plants deserve especial emphasis
on account of their economic importance either as ornamentals or
fruit producers : Chrysanthemum, Salvia, Lantana, heliotrope,
geranium, Fuchsia, Coleus, Ageratum, roses (17), egg plant, bean,
tomato, lettuce, cucumber and melons. Tobacco, when grown in a
greenhouse, shows itself to be especially attractive to the insect, but
there is no indication that this plant grown out of doors will ever be
seriously troubled by it as far north as Massachusetts. I am in-
formed by Mr. Francis Canning, head gardener at the Massachusetts
Agricultural College, that in his experience the Greenhouse Aley-
rodes has been a serious pest of Primula obconica, a plant grown
quite extensively by some florists.

ECONOMIC IMPORTANCE OF THE GREENHOUSE

ALEYRODES.

From an economic standpoint the most serious injury caused by
this species of Aleyrodes is to cucumbers and tomatoes in green-
houses. J. W. Douglas (7) speaks of cucumber and tomato plants
in England being ruined by the agency of this insect. Among the
more recent accounts of its injuries Britton (23 and 27) says : "For
eight years the most serious insect pest affecting forcing house
tomatoes at the station has been the " white fly," " mealy wing," or
plant house Aleyrodes. Were it impossible to hold the insect in
check, the crop each winter would be nearly a total failure."

Within the last two years there have been several inquiries re-
ceived at the Hatch (Mass.) Experiment Station concerning this
insect, and in at least three cases serious loss to the owners resulted

36

from its attacks. The most noteworthy of these was the total loss
of a crop of cucumbers and tomatoes of an estimated value of four
thousand dollars at Pittsfield, Mass., during the Spring of 1901.
Dr. H. T. Fernald, who visited these greenhouses about the first of
June, found every plant either dead, or nearly so, with none of the
crop ready for harvesting. Myriads of the white flies, which were
seeking in vain to obtain liquid food from the dried leaves, would
fly up in clouds when the plants were disturbed. Many other in-
stances of the destruction caused by this insect in this state and
elsewhere might be cited, but enough has been said to show that
growers of vegetables in greenhouses cannot afford to ignore it when-
ever it makes its appearance in noticeable abundance.

This species of Aleyrodes may also require treatment in con-
servatories used for private ornamental or commercial purposes. Its
importance here is largely due to its very general feeding habits.
Still another class of plants upon which injuries are frequently re-
ported are house plants, such as geraniums, heliotrope, etc.

Cucumbers and tomatoes are among the most delicate of the
plants which are liable to require treatment from the attacks of this
insect, as well as those upon which the attack is most liable to result
in serious financial loss. The treatment of these two plants has
therefore been given special prominence in this paper. The general
principles for the treatment of plants in florists' establishments are
the same as for the forcing pit, while as for the treatment of house
plants they are so comparatively unimportant that but a few words
need be said.

PREVENTIVES.

In dealing with this insect, preventives are of prime import-
ance. Owners of greenhouses which are subject to its attacks can
well afford to insure their plants against total or partial destruction
by a very small expenditure of time and money. The axiom, an
ounce of prevention is worth a pound of cure, applies here as well as
to all other human troubles. Moreover, the cure, unless one is con-
stantly on the lookout for insect pests, is in this case liable to be
applied when too late and the destruction has been already accom-
plished.

37

Greenhouses in which vegetables are grown are liable to become
infested in two ways. First, there may be weeds, such as chick-
weed or other food plants of the species, growing in the house, under
the benches or other out of the way places, and on these the insects
may live during the summer months when the house is unused, and
in the fall spread to the new crops. Again, adults which have de-
veloped out of doors may fly into the houses. This is especially
liable to occur where there are infested plants in the vicinity.

To prevent the first method of infestation, before starting a crop
in the fall, all weeds and vegetation of every kind should be re-
moved from the house. Eggs, larvae and pupae of the insect are thus
provided for. Adults may be killed by burning sulphur at the rate
of 6 oz. to iooo cu. ft. of space, or by hydrocyanic acid gas from .2
gram of Potassium cyanide (KCN) per cu. ft. of space, leaving the
house closed over night. The rate for the use of the cyanide, as
given above, is much greater than is absolutely necessary to destroy
adult Aleyrodes, in order that it may also be effective against thrips
and other pests. The effect of sulphur on other insects than Aley-
rodes is unknown to the writer.

Whenever a first crop is entirely removed, the house should be
fumigated by one of these methods before starting a new crop.
This, of course, will not be possible where there is only one house,
and the new crop is under way before the old one is removed.

To prevent the Aleyrodes from spreading into the greenhouses
from plants growing out of doors, little can be done except to see
that no infested plants are growing in the vicinity. Aleyrodes found
on strawberry plants, as will be seen later, need not be considered in
this connection. After the first few frosts in the fall there is little
danger of the insect getting into greenhouses from out of doors.

REMEDIES.

Various insecticides have been recommended and used for the
destruction of this insect. Ravenscroft (10), who was the first, as
far as I am aware, to recommend a specific treatment, says that " the
best remedy is to syringe with a solution of Calvert's Soft Soap."
Among the later recommendations are : Tobacco fumes in green-
houses, whale oil soap and kerosene emulsion out of doors (16); any
4

38

of the soap and oil washes (17) ; hydrocyanic acid gas from one
ounce of potassium cyanide for each one thousand cubic feet of
space, left in the house over night (18); common laundry soap, one
pound to eight gallons of water (23 and 27) ; nicoticide (24) ; kero-
sene and resin washes (26) ; and hydrocyanic acid gas (for adults),
using 1 oz. of potassium cyanide for each 400 cu. ft. of space, ex-
posure of nine minutes (29).

The experiments carried on at the insectary of the Massachu-
setts Agricultural College by the writer consisted of both spraying
and fumigation. The details of the experiments are first given, fol-
lowed by a discussion of results. When aphids or red spiders have
been present in noticeable abundance, the effect of the various in-
secticides on them has been incidentally noted, thereby showing the
comparative susceptibility of the Aleyrodes and other pests to the
same treatment.

Contact Insecticides.

Lemon Oil Insecticide. Mnfg. by S. I. Pollexfen, Baltimore,
Md. Cost: n to 25 cents per half pint, according to amount pur-
chased. Strength recommended: One-h'alf pint in from 4 to 12
quarts of water.

1. One-half pint in 4 quarts of water. Plants treated : To-
mato, dipped in solution 20 sec; Petunia and Eupatorium, sprayed.
Results: Plants uninjured. About 50 per cent of immature insects
(excluding eggs) killed on Petunia and Eupatorium, and about 60
per cent on tomato plant. All adults hit by spray were killed, but a
large proportion escaped by flying away. Eggs not affected.

2. One-half pint in 1 quart of water. Plant treated : Tomato,
dipped 20 sec. Results: Plant badly injured; all insects killed ex-
cept eggs.

3. One-half pint in 1.5 quarts of water. Plant treated : To-
mato, dipped 20-25 sec- Results: Plant injured, though not as
badly as in previous experiment ; all stages of the insects killed ex-
cept the eggs.

4. One-half pint in two quarts of water. Plants treated : Sin-
gle leaf of large tomato plant, dipped 10 sec. Lantana, sprayed.
Results : Tomato leaf slightly injured ; leaves of Lantana slightly

39

injured; practically all of larvae and pupae and two-thirds of adults
killed. Eggs unaffected. *

5. One-half pint in 4 quarts of water. Plants treated : To-
mato leaf nearly dead from effect of insects' attack, and another
slightly infested healthy leaf, dipped 10-12 sec. Results: Leaves
uninjured ; for some reason much more effective than in Expr. 1 ; 95
per cent of larvae and pupae killed ; eggs unaffected.

Laundry Soap (Welcome). Cost 6f to 5! cents per lb, accord-
ing to amount purchased. Strength recommended by Britton, 2
ounces to 1 gallon of water.

1. 2 ounces to 1 gallon of water. Plants treated: Salvia,
Lantana and Geranium. Results: Plants, as a whole, uninjured (a
few leaves slightly) ; about 95 per cent of larvae and young pupae
killed, and about 25 per cent of old, nearly mature pupae; about two-
thirds of adults killed ; eggs unaffected.

Stotfs Fir Tree Oil Soap. Cost 30 to 50 cents per lb., accord-
ing to amount purchased. Strength recommended : One table-
spoonful to a gallon of water.

. 1. One tablespoonful to a gallon of water. Plants treated :
Various greenhouse plants. Results : Plants uninjured ; only a few
larvae and- apparently no pupae killed; about one-half of adults killed;
eggs unaffected.

2. One ounce to a gallon of water. Plants treated : Tomato,
geranium, Petunia and Salvia. All except a geranium and Petunia
plant were syringed about one-half hour later with clear water. Re-
sults : Plants not afterward syringed with clear water were slightly
injured; about 95 per cent of larvae and young pupae, and about 25
per cent of old, nearly mature pupae killed ; a smaller proportion
killed on those plants afterward syringed with clear water. About
two-thirds of adults killed ; eggs unaffected.

Kerosene Emulsion. One-half lb. soap, 1 gallon water and 2
gallons kerosene. Usually recommended for soft bodied insects :
one part of emulsion in 9 of water.

1. One part in 6 parts of water. Plants treated : Salvia and
tomato (sprayed) ; Salvia syringed with clear water about one-half
hour later, tomato syringed the following day. Results ; Leaves of

4°

tomato plant badly spotted by the emulsion, Salvia leaves uninjured ;
practically all of larvae and about 95 per cent of pupae killed ; about
two-thirds of adults killed ; eggs unaffected.

2. One part in 3 of water. Plants treated : Arbutilon and
tomato (sprayed). The first was afterward syringed with clear water.
Results : Tomato leaves badly spotted ; practically all of larvae and
pupae and about two-thirds of adults killed ; eggs unaffected.

3. One part in 9 of water. Plants treated : Salvia and to-
mato (sprayed), also a single leaf of Salvia (dipped), afterward
washed in clear water one-half hour later. Results : On single leaf
every larva and pupa, of which there were hundreds, were killed,
and leaf showed no injury, being kept for several days in a tightly
stoppered bottle with damp cotton ; on the Salvia and tomato plants
practically all of larvae and pupae were killed, but the foliage was
slightly spotted ; two-thirds of adults killed ; eggs unaffected.

4. One part in 1 1 of water. Plants treated : Salvia, Lantana
and tomato (sprayed) ; all three syringed with clear water one-half
hour later. Results : Plants uninjured except for a few spots where
emulsion was not washed off ; practically all larvae and about 75 per
cent of pupae killed ; about two-thirds of adults killed ; eggs un-
affected.

Permbl Kerosene Soap, mnfg. by Poole & Bailey, New York
City. Cost 14 to 25 cents per lb., according to amount purchased.

1. One ounce in 1 gallon of water. The soap did not entirely
dissolve in a quart of boiling water, and the insoluble scum clogged
the nozzle somewhat. In later experiments with this material, this
difficulty was obviated by straining through cheese cloth. Plants
treated: Geraniums, heliotropes and callas. Results: Plants un-
injured ; on heliotrope practically all of larvae and pupae killed, on
geraniums results not quite as good, while on callas, where all of the
immature insects happened to be in pupal stage, only a very few
were killed. This difference in results on the different plants ap-
pears to be due to the character of the under surface of the leaves in
regard to leaf hairs ; the heliotrope, being best provided with these,
retained the liquid the longest and gave the best results ; about two-
thirds of the adults killed ; eggs unaffected.

2. Three ounces in 1 gallon of water. Plants treated : Ger-

4i

anium, Arbutelon and tomato. Results : Plants slightly injured ;
practically all of larva? and pupae and about two-thirds of adults
killed ; eggs unaffected. Plant lice on calla and red spider on
tomato apparently all killed.

3. Two ounces in 1 gallon of water. Plants treated : Salvia,
Arbutelon and tomato (sprayed). Results : Plants uninjured ; prac-
tically all of larvae and young pupae, 50 per cent of old, nearly ma-
ture pupae, and about 50 per cent of adults killed ; eggs unaffected.

Bowker's Tree Soap, mnfg. by Bowker Insecticide Co., Boston,
Mass. Cost, 6 to 8 cents per lb., according to amount purchased.

1. One ounce in 1 gallon of water. Plant treated : Tomato.
Results : Practically all of larvae and young pupae and at least 90
per cent of old, nearly mature pupae killed ; about two-thirds of
adults killed ; eggs unaffected. While the plant was practically un-
injured, on a few leaves little areas surrounding the dead larvae and
pupae were slightly affected by the application, showing that the
limit had been reached at which the solution could be used with
safety to the plant.

Good's Potash Whale Oil Soap, mnfg. by James Good, Philadel-
phia, Penn. Cost, 3^ to 5 cents per lb., according to amount pur-
chased.

1. One ounce in a gallon of water. Plant treated : Tomato-
Results : Plant uninjured ; practically all larvae and young pupae
killed ; old, nearly mature pupae apparently unaffected ; about two-
thirds of adults killed ; eggs unaffected.

2. 1 y\ ounces in 1 gallon of water. Plants treated : Tomato
(dipped and left in shade), and a badly infested leaf on another to-
mato plant (dipped 10-12 sec. and left in bright sunlight). Results :
Plant and single leaf uninjured ; practically all larvae and young
pupae and two-thirds of adults killed; about 75 per cent of nearly
mature pupae killed ; eggs unaffected.

Fumigants.*

Nicoticide, mnfg. by The Tobacco Warehousing and Trading
Co., Louisville, Kentucky. Cost, $2.50 per lb. Strength recom-
mended, one ounce to 2000 cu. ft. of space.

* Unless otherwise stated, these experiments were made in a fumigating box contain-
ing 15 cu. ft. of space, and estimated to be of about the same tightness in proportion to its
size as an ordinary well built greenhouse.

42

i. One ounce evaporated in a greenhouse containing 1888 cu.
ft. of space. Time of exposure, 6 p. m. to 7.30 a. m. Plants treated :
Various — tomato, cucumber, Salvia, Lantana, heliotrope, geranium,
etc. Results: Plants uninjured ; a small percentage of larva? and
young pupa? killed ; adults appeared as though intoxicated, and in
the course of two days, a few, perhaps 5 or 10 per cent, recovered ;
the rest finally died. No live plant lice could be found, although
they became abundant again in the course of a few weeks, showing
that a few must have escaped.

2. Rate: One ounce per 2000 cu. ft. of space. Time of ex-
posure, 6 p. m. to 7.30 a. m. Plant treated : Tomato. Results :
Plants uninjured ; adults stupified in a few minutes, but 75 per cent
fully recovered by the next morning ; larva? and pupas unaffected.

3. Rate: Two ounces per 2000 cu. ft. of space. Time of
exposure, 3 hours. Plant treated: Tomato. Results: Plant un-
injured ; adults began to drop from leaves in about 3 minutes, and
in 10 minutes none showed signs of life ; 24 hours later, 50 per cent
of adults dead ; the remainder recovered.

4. Rate: Two ounces per 2000 cu. ft. of space. Time of ex-
posure, 10 hours. Plant treated : Tomato. Results: Plant un-
injured ; adults killed ; a few larvae and young pupas kilied.

5. Rate: One ounce per 2000 cu. ft. of space. Time of ex-
posure, 23 hours. Plant treated : Cucumber. Results : Tender
leaves killed ; all adult insects killed.

6. Rate : Five ounces per 2000 cu. ft. of space. Time of ex-
posure, 7 p. M-. to 8 a. m. Plant treated : Tomato. Results : Plant
uninjured; all adults and about 50 per cent of larva? and young
pupa? killed ; eggs and nearly mature pupa? unaffected.

7. Rate: Fifty ounces per 2000 cu. ft. of space. Time of
exposure : One hour. Plant treated : Tomato. Results : Plant
uninjured ; all adults killed ; larva? and pupa? unaffected.

Buck's Best Brand Fumigating Compound, mnfg. by W. W.
Bush & Co., Ltd., London, England. Cost, $4.75 to $S.oo per lb.,
according to amount purchased. Strength recommended, one ounce
to 2275 cu. ft. of space.

1. Rate: One ounce to 2275 cu. ft. of space. Time of ex-

43

posure : 6 p. m. to 8 a. m. Plant treated : Tomato. Results :
Plant uninjured ; adults slightly disturbed.

2. Rate: Two ounces to 2275 cu. ft. of space. Time, 3
hours. Plants treated : Tomato and geranium. Results : Plants
uninjured; adults began to drop from leaves in about 2 or j min-
utes ; in the course of 3 days, however, all but about 5 per cent fully
recovered.

3. Rate: One ounce to 2275 cu. ft. of space. Time of ex-
posure, 22 hours. Plant: Tomato. Results: Plant uninjured;
a few adults showed signs of life when door of fumigating box was
opened, but at the end of the third day after fumigation all adults
were dead ; a few larvae and pupae killed.

4. Rate: Two ounces to 2275 cu. ft. of space. Time of ex-
posure, 6 p. m. to 8 a. m. Plant treated : Tomato. Results : Plant
uninjured ; a few larvae and young pupae and 95 per cent of adults
killed.

Thripscide, mnfg. by E. H. Hunt, 76 Wabash Ave., Chicago, 111.
Cost, 16^- to 25 cents per lb., according to amount purchased.
Strength recommended, One lb. for an ordinary 100 foot greenhouse
for' a light fumigation, or 5 lbs. for an ordinary 300 foot greenhouse
for a heavy fumigation.

1. One-fourth lb. in a greenhouse 16 feet long containing 1700
cu. ft. of space. Time of exposure, 7 p. m. to 8 a. m. Plants :
Hibiscus and Eupatorium. Results: Plants uninjured; adult
Aleyrodes stupified, but all recovered ; apparently all plant lice
killed.

Carbon Bisuljid. Ordinary commercial carbon bisulfid was
used in these experiments. This costs about 25 cents per lb. Furaa
Bisulfid is manufactured by E. R. Taylor, Penn Yan, N. Y., and
costs 10 cents per lb.

1. Rate : One lb. per 1000 cu. ft. of space. CS3 was poured
on to cotton in this and later experiments. Time of exposure, one
hour ; sun shining, 8.45 a. m. Plant treated : Tomato. Results :
Adults began to drop from leaves in about 25 minutes, and showed
but slight signs of life when fumigating box was opened ; all recov-
ered in course of 2 hours ; plant uninjured.

44

2. Rate: Two lbs. per iooo cu. ft. of space. Time of ex-
posure, i hr., 15 min. ; sun shining. Plant treated: Tomato. Re-
sults: Adults began to drop from leaves in about 15 minutes; ap-
parently dead when box was opened, but all recovered in the course
of three hours ; plant uninjured.

3. Rate: Two lbs. per 1000 cu. ft. of space. Time of ex-
posure, 6 p. m. to 8 a. m. Plant treated : Tomato. Results :
Adults apparently killed, but about 25 per cent recovered in the
course of two days ; plant uninjured.

4. Rate: One and one-half lbs. per 1000 cu. ft. of space.
Time of exposure, 7 p. m. to 8 a. m. Plant treated : Tomato. Re-
sults : Plant uninjured ; adults apparently killed, but in the course
of two days about 20 per cent, recovered; about 10 per cent of
young larvae killed.

5. Rate: Five lbs. per 1000 cu. ft. of space. Time of expo-
sure, 7 p. m. to 7.30 a. m. Plant treated: Tomato. Results: Plant
uninjured ; adults all killed ; 50 per cent of larvae and pupa; killed ;
eggs unaffected.

Hydrocyanic Acid Gas. Cost of Potassium cyanide (KCN)
about 50 cents per lb. ; of Sulphuric acid (H2S04) from 2^ to 10
cents per lb.

1. Rate: .1 gm. KCN per cu. ft. of space in a practically air
tight fumigating box containing 30 cu. ft. of space. Time of expo-
sure, 10 minutes, daylight. Temperature, 400 F. Plant treated:
Tomato. Results : Tender leaves killed and older leaves injured
at the tips ; all larvae were killed ; no pupae or adults present ; eggs
unaffected.

2. Rate: .05 gm. KCN per cu. ft. of space in the same box
that was used in Expr. 1. Time of exposure, 10 minutes, daylight.
Temperature, 48° F. Plant treated : Tomato. Results : Plant
very slightly injured; very few larvae killed; eggs unaffected ; no
other stages present.

3. Rate: .0016 gm. KCN per cu. ft.: 3 gr. in greenhouse
containing 1888 cu. ft. of space. Time of exposure, 6 p. m. to 8
a. m. Plants treated : Various greenhouse plants. Results : Plants
uninjured; adults were found to be stupified and fluttered about on
the benches ; all but about 5 per cent, recovered during the day.

45

4. Rate: .01 gm. KCN per cu. ft. of space (see footnote page
41). Time of exposure, 30 minutes, daylight. Plant treated : To-
mato. Results : Extreme tips of some of larger leaves showed in-
jury ; growing tips of stems and the tenderest leaves uninjured ; all
adults killed, many being left suspended to the under surface of the
leaves by their rostral seta; and tarsal claws ^ none ever showed signs
of life after box was opened.

5. Rate : .01 gm. KCN per cu. ft. of space. Time of expo-
sure, 30 minutes, cloudy day. Plant treated : Tomato nearly dead
from the effects of the insects' attacks, and a sprig of Eupatorium
infested with plant lice. Results : After a few days the young and
tender leaves of the plant appeared injured, [perhaps due to the
fumigation ; all adults killed ; eggs, larvae and pupae apparently un-
affected. Plant lice stupified, but recovered during the day after the
fumigation.

6. Rate : .007 gm. KCN per cu. ft. of space. Time of expo-
sure, 30 minutes, after sunset. Plants treated : Tomato and cucum-
ber. Results : Plants not injured in the least, the cucumber plant
being in full bloom two days later ; many of the adult Aleyrodes
continued to show signs of life during the next day, but 36 hours
later all were dead ; eggs, larvae and pupae unaffected.

7. Rate : .01 gm. KCN per cu. ft. of space. Time of expo-
sure, 25 minutes, cloudy day. Plants treated : Tomato and cucum-
ber used in experiment 6 and another tomato plant never before
treated. Results : Plants uninjured, even blossoms and flower buds
of cucumber escaping; all adults killed; eggs, larvae and pupae un-
affected.

8. Rate: .01 gm. KCN per cu. ft. of space. Time of expo-
sure, 20 minutes, cloudy day. Plants treated: Tomato and cucum-
ber used in experiments 6 and 7, and a cucumber plant never before
treated. Results : One leaf of tomato slightly injured, plants other-
wise unaffected ; all adults killed ; eggs, larvae and pupae unaffected.

9. Rate : .01 gm. KCN per cu. ft. of space. Time of expo-
sure, 20 minutes, after sunset. Plants treated; Tomato, badly in-
fested with Aleyrodes and red spider, and sprig of Eupatorium in-
fested with plant lice and Aleyrodes. Results: Plants uninjured ;
all adult Aleyrodes killed; red spicier unaffected; plant lice all stupi-

46

fied and apparently killed, but all recovered in the course of 12
hours.

10. Rate : .01 gm. KCN per cu. ft. of space. Time of expo-
sure, 15 minutes, cloudy afternoon. Plants treated: Cucumber
and tomato. Results: Plants uninjured; about 50 per cent of
adults were dead when box was opened, but the remainder showed
signs of life, and some of these finally fully recovered.

1 1. Rate : .005 gm. KCN per cu. ft. of space. Time of expo-
sure, 30 minutes, after sunset. Plants treated : Cucumber and to-
mato, and a sprig of Eupatorium infested with plant lice. Results :
Plants uninjured ; all adult Aleyrodes killed, though many showed
slight signs of life for a few honrs. Plant lice stupified, but recov-
ered within twelve hours.

12. Rate : .005 gm. KCN per cu. ft. of space. Time of expo-
sure, after sunset, 6 to 9 p. m. Plants treated: Tomato and cu-
cumber, the latter badly infested with red spider. Results : Plants
uninjured; all adults killed, a few showing signs of life for a few
hours ; two-thirds of larvae and pupae killed ; eggs unaffected.

13. Rate : .007 gm. KCN percu. ft.-of space in a loose green-
house containing 1700 cu. ft. of space. Time of exposure, after
sunset, 6.45 to 9.45 p. M. Plants : Tomatoes, with many of leaves
nearly dead from results of insects' attack, and cucumber infested
with red spider. Results : Plants uninjured ; all adult Aleyrodes
killed ; 90 per cent of larvae and young pupae killed, and 2 or 3 per
cent of old, nearly mature pupae killed ; eggs unaffected ; red spider
unaffected.

14. Rate ; .01 gm. KCN per cu. ft. of space in same green-
house used in last experiment. Time of exposure, after sunset, 6.45
to 9.45 p. m. Results : Plant uninjured ; all of adults, all larvae
and young pupae and about 75 per cent of old, nearly mature pupae
killed ; eggs apparently unaffected.

15. Rate: .02 gm. KCN per cu. ft. of space (in fumigating
box). Time of exposure, after sunset, 7.30 to 10.30 p. m. Plants :
Tomato and cucumber, the former being small plants, showing the
effects of improper nutrition, the latter being badly infested with red
spider. Results : The smaller and less vigorous of the two tomato
plants used showed slight injury to some of the tenderest leaves ; the

47

other plants showed no injury ; on the contrary, a cheek cucumber
plant, which was practically a duplicate of the one fumigated in re-
spect to size, vigor and infestation, was dead two days later, while
the fumigated plant rapidly improved in health. Of the Aleyrodes,
all the adults, larva? and pupa? were killed; eggs apparently killed.
The red spiders on the cucumber plant were all killed.

DISCUSSION OF RESULTS.

Contact Insecticides. Few contact insecticides were used, but
these were varied enough to indicate that in general they are destruc-
tive to larva? and young pupa?, and more or less so to nearly mature
pupa?. They proved useless, however, for the eggs, and not practical
for use against the adults of this insect. The statements of those
who have had practical experience with Aleyrodes in greenhouses
shows that syringing the plants does not seem to lessen the numbers
of the adults. In these experiments especial attention was given to
the destruction of the adults, but not more than two-thirds of them
were killed in any case. At the most, one could not expect to kill
more than one-half the adults by an ordinary syringing, which would
be sufficient to destroy 90 per cent of the young. The most expert
growers of tomatoes in greenhouses tell us that these plants should
not be syringed except when it is absolutely necessary, for a damp
atmosphere promotes rot and interferes with pollination. It is de-
sirable, therefore, to depend as much as possible on fumigation for
the control of the greenhouse Aleyrodes, at least on those plants
which are liable to be injured in any way by the application of a
contact insecticide. These latter, however, may be used to advan-
tage in many cases.

The following comments will aid in showing the relative value
of each of the contact insecticides used in the experiments:

Lemon Oil Insecticide. — The cost of this material precludes its
use on a large scale, but it is easy to prepare, has a not unpleasant
odor, and the experiments show it to be effective on the larva? and
pupa? when the plant is dipped in a solution of one-half pint in one
gallon of water. Chiefly commendable for the treatment of house
plants.

Laundry Soap. — Cheap; fairly effective on the insects when

48

used at the rate of two ounces in a gallon of water ; liable to injure
foliage slightly.

Stott's Fir Tree Oil Soap. — Altogether too expensive for use
against the insect, even though it were effective.

Kerosene Emulsion. — Cheap ; troublesome to prepare ; effec-
tive as an insecticide, but requires syringing the leaves afterward to
prevent injury, at least in greenhouses when used on tomato plants.

Permol Kerosene Soap.- — Too expensive ; fairly effective when
used at the rate of two ounces in a gallon of water.

Bowker's Tree Soap. — Cheap ; most effective of all contact in-
secticides used in these experiments ; has advantage over the others
in requiring but one ounce in a gallon of water, making it less liable
to interfere with the respiration of the plants by forming a film over
the surface. While the odor of the soap itself is unpleasant, it is not
very objectionable when used in such a weak solution.

Good's Potash Whale Oil Soap. — Cheaper by weight than Bow-
ker's Tree Soap, but when an amount is used which equals the latter
as an insecticide, the cost is about the same for the two soaps. One
and one-third ounces per gallon of Good's Soap is nearly as efficient
for use on the insect as one ounce of Bowker's Soap. More efficient
than laundry soap or Permol Kerosene Soap. The same may be
said of the odor as of Bowker's Tree Soap.

Fumigants. The resistance of the greenhouse Aleyrodes to the
action of tobacco fumes is well known. The young, as might be ex-
pected from their mode of life, are entirely unaffected, while the adults
are only temporarily stupified by the same treatment with tobacco
that destroys most of the plant lice. The experiments with Nicoto-
cide and Thripscide show that these, also, are much more effective
on plant lice than on the adult Aleyrodes, while the experiments with
hydrocyanic acid gas show the remarkable susceptibility of adult
Aleyrodes to its action as compared with plant lice, the latter being
only stupified in the same atmosphere which was fatal to the former
Small amounts of the fumigants with a long exposure seem to be
more effective on the larva? and pupae than large amounts with short
exposures.

Nicoticide. — The difference in the results of experiments i and
2 may be accounted for by the fact that in experiment i the tomato

49

plant was a small one, and the pot rested on the floor of the fumi-
gating box. It was observed that the fumes were very thin near
the bottom of the box, and in later experiments the pots were set on
a board raised about a foot from the floor. This insecticide is in-
expensive to use, and some who have had experience with it against
the insect claim that it is ineffective even for the adult. My own
experiments show that one or two ounces to 2000 cu. ft. of space,
according to the tightness of the house, should result in killing the
adults. The cost of fumigating a house containing 20,000 cu. ft. of
space at one ounce per 2000 cu. ft. of space would be $1.56. Ac-
cording to experiment 6 it would cost five times this amount to kill
one-half of the larvae and pupa; in a house of the same size. There
are many insecticides on the market which are used as fumigants
which have tobacco extracts as the active principle. Nicoticide may
be considered as a fairly representative type of the best of this class.

Bush's Best Brand Fumigating Compound. — Nearly as effective,
ounce for ounce, as Nicoticide, but much more expensive.

Thripscide. — The single experiment indicates that this material
is no more effective on the adult Aleyrodes than ordinary tobacco.

•Carbon bisulfid. — Long exposure necessary for good results ;
too expensive for ordinary use. In a greenhouse containing 20,000
cu. ft. of space, the adults and a small percentage of the young could
be destroyed for $6.00, and the adults and about 50 percent of the
immature insects for about $10.00.

Hydrocyanic Acid Gas. — Safe to use only by careful and intelli-
gent persons who appreciate its properties. Far cheaper and more
effective than any other fumigant known for use against the green-
house Aleyrodes. Cost for a house containing 20,000 cu. ft. of
space, using KCN at rate of .01 gm. per cu. ft., between 25 and 30
cents for each fumigation. Even plants as tender as tomatoes can
be fumigated without injury with the extremely small amounts of the
cyanide required.

RECOMMENDATIONS.

Fumigation. Bearing in mind the cost, labor involved and effect
on plants and Aleyrodes of the various materials used in these ex-
periments, hydrocyanic acid gas is pre-eminently the most deserving

5°

of recommendation. From the fact that no two greenhouses are ex-
actly alike in regard to tightness, it is impossible to give specific
directions to cover all cases, but it is safe to recommend the use of
an amount of potassium cyanide varying according to the tightness
of the house from .007 to .01 gm. per cu. ft. of space, for three hours'
time after sunset. No injury to tomato or cucumber plants would
result from using the larger amount, but it is believed that the smal-
ler amount in a tight house will prove as effective as the larger
amount in a loose house. Even .02 gm. per cu. ft. of space, as shown
by Experiment 15, can be used without injury to cucumber and to-
mato plants in a reasonable state of vigor. For tender plants like
tomatoes it is advisable to keep well t>elow the danger limit of the
tenderest leaves and use the potassium cyanide at rates not greater
than .01 gm. per cu. ft. of space.

It is not within the scope of the present paper to give the minute
details of fumigating with this most powerful of all known destroyers
of animal life.

Johnson's recent work, entitled " Fumigation Methods," is in-
dispensable to anyone who has occasion to use hydrocyanic acid gas
as an insecticide. This book can be purchased for $1.50 of the pub-
lishers, Orange Judd Company, 52 and 54 Lafayette Place, New
York City. To those interested in greenhouse fumigation the fol-
lowing pages are especially recommended : 9-1 1, 118, 124-146.

Attention is here called to the fact that ventilation of the house
is necessary after the expiration of the three hours of treatment, and
that this is practicable only under such circumstances that no injury to
the plants will result from lowering the temperature of the house.
The amount of artificial heat that can be supplied, and the outside
temperature must be taken into consideration. Ventilation need not
be very prolonged nor very thorough immediately following the ex-
posure ; a few side ventilators opened (from the outside) for a quar-
ter of an hour will be sufficient. A more thorough ventilation should
be given before one ventures to inhale air while in the house.

In order to be most successful, fumigation should be in accord-
ance with a knowledge of the life history of the insect. A single
fumigation at the rates recommended will destroy all of the insects
except those in the egg and some of those in the late pupal stage.
.It is desirable to fumigate again two weeks later, thus killing the

5'

few adults that have emerged in the meantime, and which at the time
of the first fumigation were in the late pupal stage, and all the larva;
which resulted from the eggs present at the first fumigation. The
adults killed by the second fumigation will have deposited a few
eggs, the larva; from which will be in the first, second and third in-
stars two weeks after the second fumigation, and a third treatment
at this time will destroy them. Theoretically the house will now be
entirely freed from the pest. The writer's tests indicate that this is
possible, but even though it is not actually accomplished, no trouble
need be expected from the insect for many weeks. In badly infested
houses even a single fumigation will so reduce the numbers of the
insect as to prevent furthur injury for at least a month, and an occa-
sional fumigation in slightly infested greenhouses will control the
pest so as to prevent the possibility of its increasing to an injurious
extent.

Fumigation Combined with Syringing. If for any reason it seems
undesirable to use hydrocyanic acid gas, a systematic treatment with
a fumigant and a contact insecticide should result in effectually con-
trolling the insects. The first step in the process should be to pre-
vent, the enormous production of eggs by ridding the house of the
adults. For this purpose, Nicoticide is the most effective and cheap-
est material, aside from hydrocyanic acid gas, which was used in the
foregoing experiments. Other similar preparations may be found to
be equally good for the purpose. For each two thousand cubic feet
of space from one to two ounces of Nicoticide should be used, ac-
cording to the tightness of the house. This fumigation should last
all one night, and be followed the next day by a thorough syringing
of the plants with a contact insecticide, the most desirable used by
the writer being P>owker's Tree Soap, at the rate of one ounce in a
gallon of water. One such combined treatment should so reduce the
numbers of Aleyrodes in a greenhouse that further treatment would
be unnecessary for several weeks. If it is desired to continue this
treatment further, the house should be fumigated as before, one week
after and then sprayed as before, two weeks after the first treatment.

52

BIBLIOGRAPHY OF ALEYRODES VAPORARIORUM

i. Westwood, Gardeners' Chronicle, p. 852 (1856).

2. Signoret, Ann. de la Soc. Ent. Fr., 1868, p. 3S7 (1867).

3. Frauenfeld, Verh. derZool. Bot. Gesellsch., Wien, p. 798 (1867).

4. Packard, Agric. of Mass., 1 869-1 870, p. 262 (1S70).

5. Packard, Am. Nat., vol. iv, p. 6S6 (187 1).*

6. Packard, Guide to Study of Insects, p. 526 (1883).

7. Douglas, Ent. Mo. Mag., vol. xxiii, pp. 165-166 (18S6).

8. Comstock, Second Ann. Rept. Cornell Expr. Sta., p. 20 (1889).
9. -, Dictionary of Gardening, vol. iv, p. 52 (1889).

10. Ravenscroft, Tomato Culture, L. Upcott Gill, Pub., p. 77

(1S90).

11. Riley and Howard, Insect Life, vol. ii, p. 339 (1890).

12. Weir. Insect Life, vol. iii, p. 394 (1S91).

13. Bailey, Bui. Cornell Expr. Sta., No. 28, p. 58 (1S91),

14. Riley and Howard, Insect Life, vol. iv, p. 345 (1892).

15. Riley and Howard, Insect Life, vol. v, p. 17 (1893). *

16. Britton, 19th Rept. Conn. Expr. Sta., pp. 203-204 (1895).

17. Davis, Miscellaneous Bui., (Spec. Bui. 2), Mich. Expr. Sta.,

pp. 22-24 (1896).

18. Fisher, Am. Gardening, vol. xix, No. 201, p. 741 (1898).

19. Quaintance, Tech. Ser. No. 8, U. S. Dept. Agric, Div. Ent.,

pp. 16, 39 (1900).

20. Britton, Report Conn. Expr. Station, pp. 311-312 (1900).

21. Rane, Bui. N. H. College Expr. Sta. No. 84, p. 67 (1901).

22. T , Florists' Exchange, vol. xiv, No. 33, p. 914 (1902).

23. Britton, Bui. Conn. Expr. Sta. No. 140, pp. 1-17 (1902).

24. Beal, Bui. 111. Expr. Sta., No. 81, pp. 512, 522 (1902).

25. Morrill, Can. Ent., vol. xxxv, No. 2, pp. 25-26 (1903).

26. Chittenden, Am. Florist, vol. xx, No. 774, p. 3S6 C1903).

27. Britton, 2d Ann. Rept. Conn. State Ent., pp. 148-163 (1903).

28. Morrill, Psyche, vol. x, No. 322, pp. 80-82 (1903).

29. Weed and Conradi, Bui. N. H. College Expr. Sta. No. 100

Pj>47-52 (T9°3)-

* A. vaporariorum in part.

\

53

The Strawberry Aleyrodes, Aleyrodes packardi

Morrill.

This well known and widely spread insect was for many years
believed to be identical with the common greenhouse Aleyrodes.
Packard, who first mentioned this insect in the American Naturalist
(i) as occurring in large numbers on strawberry plants at Amherst,
Mass., referred it to the species " vaporarium " (vaporariorum). This
was a very excusable error, as we are justified in separating the two
species only after a critical study of all the stages of each, laying
special stress upon the range of variation in their structure. Since
this first mention the insect* has been reported several times as oc-
curring on strawberries (in such regions) in Ohio (7), Kentucky, (4,
5, 8, 10, 12), Southeastern New York (9) and Connecticut (10, 12).

The food plants of the species, as far as known, are very lim-
ited. Besides the strawberry, pupae of this insect have been found
in small numbers on ash, spiraea and camperdown elms.

I have recently described (11) all the stages of the strawberry
Aleyrodes, and have briefly tabulated the differences between it and
the greenhouse Aleyrodes. I will here consider more in detail the
points wherein the former species differs from the latter.

Egg. Length varies from .23 to .24 mm. ; greatest width from
.08 to .094 mm.

First Instar (Plate VI, Fig. 33). Only sixteen pairs of mar-
ginal spines are present in this instar, there being none to correspond
in position to the seventh and ninth pairs of A. vaporariorum. The
average comparative lengths of these spines in A. packardi are about
as follows :

* I have positively identified specimens from Kentucky believed to be A. vaporariorum
as belonging to A. packardi. Ouaintance (9) has considered specimens from New York
state on strawberries which were too poor for positive identification as "much like A.
vaporariorum.'1'' My experience with strawberry plants growing in a plant house thickly
infested with A. vaporariorum goes to show that while the larva of this species will grow
to maturity on that plant if transferred in the first instar, the adults show no liking for it,
and are almost never observed even resting on its leaves. On such a plant I once found a
very few pupa cases, which were, however, too poor for identification. It is possible that
they were A. vaporariorum. I have never found an A. vaporariorum on a strawberry
plant out of doors. On the whole, therefore, I feel justified in considering all Aleyrodes
which have been reported as occurring on strawberries in this country and taken for
A. vaporariorum as belonging to the species packardi.

5

54

3 4 5 o 7 b 9 io ii 12 13 14

» — > — » — , > > > ■

6|' 6f 7 ' 5 ' 5 4* 4 3 2 2' 2 2 3
15 16

11'

4' 24"

The true lengths in mm. of these spines may be obtained by
multiplying the lower figures by .0035.

On the cephalic region of the dorsum I have been unable to
distinguish a pair of spines to correspond with those in A vapora-
riorum as well as in the later instars of A. packardi itself.

The length of the body in the first instar varies from .29 to
.35 mm., the greatest width from .16 to .18 mm.

Second and Third Instars. In these stages the only difference
in structure observed between the two species is in the comparative
lengths of the first and third pairs of dorsal spines, the second pair
in both species being minute. In A. packardi all three pairs are in-
variably minute.

The length of the body in the second instar varies from .41 to
.45 mm., the greatest width from .21 to .26 mm.

The length of the body in the third instar varies from .56 to
.62 mm., the greatest width from .32 to .38 mm.

Pupa. The most marked differences between the two species
are in the pupal stage, although at a casual glance they appear alike.
In A. packardi all three pairs of dorsal spines are minute, while in
A. vaporariorum the third pair is very variable in its degree of de-
velopment, as already described. In the former there is only a
double sub-marginal series of wax rods present, none arising farther
up on the dorsum. The rods of the outer series, unlike those of A.
vaporariorum, are very variable in length, in mature pupae being
usually more than one-half the width of the body in length. The
number of rods in the outer series averages much greater in A. pack-
ardi than in A. vaporariorum, there being from about sixty to one
hundred present in the former and from about fifty to seventy-five
in the latter species. The inner sub-marginal series of rods also
differs in number in the two species, there being eighteen or twenty
in A. packardi and but ten or twelve in A- vaporariorum. In both
species these (the inner-submarginal series) are usually directed up-,
ward and curved inward over the dorsum of the body, but the varia-
tion in their place of origin is much less in A. packardi, none having

55

been observed to arise farther mesad from the outer series than the
width of their bases.

The length of the pupa of A. packardi varies from .748 to
.88 mm., the greatest width from .407 to .54 mm.

Adult. The following formula shows about the average propor-
tionate lengths of the segments of the antennae in the female : 2 —
5 — io| — 3 — 4 — 3^ — 3^. The segments of the antennae of the males
are slightly smaller, but show about the same proportion.

The two basal segments seem to be of little systematic import-
ance, as they can rarely be accurately measured owing to the fact
that in specimens mounted in balsam their long axis is seldom perpen-
dicular to the line of vision. I have therefore considered 2 — 5 their
proportionate length in both species, these numbers being approxi-
mately correct. In A. packardi the following formula indicates the
range of variation in length observed in each segment, omitting the
two basal ones: (10-12)— 3— (4-4|)— (3-4)— (3-5).

For convenience of comparison the following formula, repre-
senting the range of variation observed in the same segments of A.
vaporariortim, is again given : (9-1 1) — 4 — (5-6) — (4-5) — (4-5)-

The length of the fore wing of A. packardi is about 1.16 mm.,
the breadth about .48 mm.

The length of the body of the adult female varies from 1.15 to
1.20 mm. ; average length of the male about .90 mm.

NOTES ON LIFE HISTORY AND INFLUENCE OF

WEATHER.

This species evidently passes the winter entirely in the egg
stage. Here in Amherst the adults become extremely abundant
every fall, and hundreds, or even thousands, of eggs are deposited on
nearly every leaf. The adults are very hardy, and egg laying may
continue up to the middle of November, the larvae and pupae having
all been killed by the frosts some weeks earlier.

In the spring a few consecutive warm days in March cause
many of the eggs to hatch. The young larvae, however, being in the
most delicate stage, will, as a rule, succumb to the frosts up to the
first of April, or even later. In the spring of 1902 all the eggs had
hatched by the end of the first week in May. The first adults
emerged about the middle of May, and from this time until Novem-

56

ber all stages could be found on the leaves. In the spring of 1903,
two adults, recently emerged from their pupa cases, were found on
May 4, but it was over a week later before the adults became at all
common. In some seasons the strawberry Aleyrodes will be much
more abundant than in others, owing to the varying weather condi-
tions. A cold March, unfavorable to the development of the eggs,
will delay the time of hatching until the larvee will be less liable to
be cut off by the spring frosts, while a couple of weeks of warm
weather in March, followed by a cold snap, will cause a large num-
ber of the eggs to hatch and the resulting larvae to die, and conse-
quently comparatively few adults will result from the wintering over

eggs.

Of those larva? which are fortunate enough to escape destruc-
tion by the frosts only a small proportion. reach maturity, as most of
the old leaves upon which they feed and from which they are unable
to move, are dead before the insects reach maturity. Considering,
however, the enormous number of eggs which are deposited upon
the leaves in the fall it is evident that if only one adult results from
each one thousand eggs, the number left to continue the species will
still be large.

ECONOMIC IMPORTANCE OF THE STRAWBERRY

ALEYRODES.

The only serious injury thus far reported as resulting from the
attacks of this insect has been from the state of New York. The
following quotation from Slingerland (9) shows the amount of injury
the insect may do :

" In the fall of 1897, and again in July, 1900, we received speci-
mens of strawberry leaves which were seriously infested with a pe-
culiar scale-like insect. The first specimens came from Sparkill,
Rockland county, N. Y., and those last year from Rossville, Staten
Island, or not far from the first. The following statements from
letters of our correspondents will describe the insects' work:

' Our strawberry plants are full of very small white flies. They
seem to suck the sap out of the leaves ; they are on the under side
of the leaves, and when disturbed fly away. The leaves turn black

57

on the outer edges, where they are infested badly, and some plants
are nearly or quite killed by them. Certain varieties are more in-
fested than others.'

' The leaves I enclose were taken from plants set this spring,
which have been attacked by small white insects on the under side
of the leaves,' writes our Rossville correspondent in July, 1900.
' When I touched the plants, the flies, not larger than a grain of salt,
but perfectly white, would rise up by the thousands in clouds. The
plants started off vigorously with large healthy runners. Finally, I
noticed that the plants began to look dead, leaves began to die and
the runners began to wilt and dry up. Some of the plants are dead.
The patches that were in bearing were also found to be badly in-
fested later in the season. While picking the fruit the upper sides
of the leaves seemed glossy like varnish, and the pickers remarked
that their hands were covered with stickiness. Later on the plants
had a black smutty appearance. These plants were very vigorous,
but now the greater part of them have turned brown and died out
entirely.' "

From my own observations I consider that the greatest injury
to strawberries by these insects is in the late summer and early fall.
The adults and immature insects are then very abundant, and though
comparatively few leaves are actually killed, the plants are percepti-
bly weakened and fail to develop good vigorous crowns and roots.
As a consequence, the following spring the new growth is retarded
and the fruit yield reduced.

In the springs of 1902 and 1903 the insects did not become
sufficiently abundant previous to the fruit harvesting season to do
any appreciable damage to the fruiting vines. It seems, therefore,
that whenever the insects become of sufficient importance to require
treatment, that such treatment should be applied so as to prevent
the injury which takes place in late summer and early fall.

TREATMENT.

In considering the treatment of the strawberry Aleyrodes we are
confronted by two problems : How to prevent the insects from es-
tablishing themselves on newly set plants, and how to deal with them
when once they are present in large numbers.

53

The amount of labor, time and money which it is expedient to
spend on the treatment of the insect in strawberry fields depends in
a large measure upon its abundance and the past experience of the
owner. Each case must be judged on its own merits. In most
cases it probably will not pay to treat the insect, but in cases as
serious as those referred to from New York it demands considerable
attention. When experience shows that treatment is advisable, the
following recommendations will be of use.

Preventives.

i. Strawberry plants known to be infested should not be in-
troduced into uninfested regions, unless it be at a season when
the recommendations given under remedies for nursery plants are
applicable.

2. In locating new fields those not adjacent to old infested
fields are to be preferred. As the natural spread of this insect is
clue largely to winds rather than to its powers of flight the value of
thus locating a new strawberry field is obvious. It has been the ex-
perience of the writer that the^Aleyrodesoccur in greatest abundance
on those fields which are nearest to the old infested ones.

3. Plants for propagation should be grown for that purpose
alone in a nursery isolated from the fruiting beds. This method not
only acts as a preventive against the insect but is in accordance
with the best methods of strawberry culture.

4. If plants are chosen from an infested field for setting pur-
poses they should be taken, when possible, from the least infested
parts of the field.

5. Plow under the old infested fields as soon as the fruit is
harvested.

Remedies.

1. In the Nursery. When plants for setting purposes are
grown in separate plots or nurseries, they should be first thoroughly
treated there, as in such cases the next year's fruiting plants are con-
fined to a comparatively small area, and the trouble and expense of
treatment is proportionately less. The number of Aleyrodes on such
plots can be reduced to a minimum by pinching off and removing in

59

the spring the old leaves which have lived through the winter. Until
the adult, or winged form, appears in the spring, shortly after the
first of May, the insects are confined wholly to these leaves, and
there is no chance for them to spread to the new spring growth.
The old leaves are usually weak and spotted with rust, and the plants
are, on the whole, benefitted rather than injured by their removal.
They lie close to the ground, and their dark green color contrasts
sharply with the bright light green of the newer leaves. The ex-
pense of removing these old leaves will be lessened if it is clone in
connection with weeding, but in any case the expense will be trivial.
Experienced strawberry growers advocate setting new fields as early
in the spring as the ground can be worked to advantage, perhaps
between the middle of April and the first of May in this latitude.
Plants taken from the nursery for setting purposes will be freed from
this insect in direct proportion to the thoroughness with which the
old wintered over leaves are removed.

If the adults are allowed to emerge and become scattered over
the leaves so that a second brood of larvae appears, the insect should
be held in check in the nursery by an occasional thorough spraying
with kerosene emulsion (one part in ten of water) or whale oil soap
(1.5- ounces Bowker's Tree Soap or 2 ounces Good's Potash Whale
Oil Soap to a gallon of water). The spray, in order to be effec-

tive, should reach the under surface of ths leaves, and for this pur-
pose an underspray nozzle can be used to advantage.

During the first week in May, 1902, the writer tried fumigation
of growing strawberry plants in order to determine whether nurseries
can be freed from the insect in this manner. While the results are
apparently not of much practical value they are nevertheless inter-
esting, as showing the effect of potassium cyanide and carbon bisulfid
on the larvae of Aleyrodes and on growing strawberry plants.

In the experiments with carbon bisulfid, the liquid was poured
on to cotton and placed near the plant to be fumigated ; a bell jar
with dirt tightly packed around its base was used as a cover. In
the experiments with Potassium cyanide a large dry goods box, made
practically air tight with putty, was first placed over the plant to be
treated and made to set squarely on the ground, with loose soil all
around in position to be quickly packed against the sides. The box
was raised from one end and the potassium cyanide, wrapped in

6o

filter paper, was dropped into the glass vessel containing the freshly
mixed sulphuric acid and water. The box was quickly dropped into
position and made air tight around the base with dirt.

Experiments With Carbon Bisulfid.

No.

3

4

5
6

7
8

9

io

1 1

12

13

14

*5

Rate.

10 gms. per cu. ft.

15 "
10

iS "

20 " "

25 "

30 " "

15 "

20 " "

10 " "

10 "

.2 gm. per cu. ft.

Time.

J5

minutes.

J5

(i

30

a

3°

Ct

3°

it

3°

a

3°

ti

hour.

5 minutes.

10

J5

10

Results.

\ Insects unaffected.
^ Plant uninjured.

Insects unaffected.

Plant uninjured.
{ Insects unaffected.
( Plant uninjured.
( Insects unaffected.
( Plant uninjured,
j 1 or 2 p. c. insects killed.
I Plant uninjured.
\ 80 per ct. insects killed.
| Plant slightly injured.
j Insects all killed.

Plant killed.

Insects all killed.

Plant badly injured,
f Insects all killed.
\ Plant killed.
\ 20 per ct. insects killed.
| Plant uninjured.
f 25 per ct. insects killed.
( Plant uninjured.
\ Insects unaffected.
( Plant uninjured.
\ 60 per ct. insects killed.
I Plant uninjured.
f Insects all killed.
^ Plant slightly injured.
\ 75 per ct. insects killed.
^ Plant slightly injured.

I

The eggs having all hatched at the time these experiments were
made, the effect on that stage could not be observed. The conclu-
sion to be drawn from the above is that for short exposures there is

6i

very little difference between the strength of these gases required to
kill the young of this insect and the strength that will injure the
•growing plant. Experiments numbers 10 and n seem to indicate
that as regards the killing of the insect without injuring the straw-
berry plant, a small amount of carbon bisulfid for a long time is
better than a large amount for a short time.

Tests of the use of Potassium cyanide for the strawberry root
louse by Powell and Sanderson of Delaware, and also by Johnson
(Fumigation Methods, p. 14S), show that two-tenths of a gram of
KCN' is satisfactory when properly handled. The plants are taken
from the nursery, freed as much as possible from dirt and moisture,
and loosely packed upon trays in the fumigating box or room.

To determine the effect of this treatment on the eggs of the
strawberry Aleyrodes, a large number of strawberry leaves with hun-
dreds of the insects' eggs attached to their under surfaces were treated
in a fumigating box, using .2 gm. KCN for twenty minutes. The
result was that none of the eggs hatched, although conditions in the
laboratory were favorable, as eggs on untreated leaves invariably
hatched under similar conditions. This shows that when strawberry
plants are treated according to the above method for the root louse,
the treatment will also be effective for the Aleyrodes when present.

2. In the Field. This is the most difficult place to treat the
strawberry Aleyrodes. If they become very numerous in the fruit-
ing or newly set beds, and require treatment, the best that can be
recommended at present is to spray with an underspray nozzle as
described for the nursery.

BIBLIOGRAPHY OF ALEYRODES PACKARDI

MORRILL.

1. Aleurodes vaporarium, Packard, Am. Nat., vol. iv, p. 686

(i87i>

2. Aleurodes vaporarium, Packard, Guide to Study of Insects, p.

712 (1883).

3. Aleurodes sp., Forbes, 13th Ann. Rept. Insects of 111., p. 98

and I (1883).*

* Page 9S footnote, and page 1 Addenda and Corrigenda, " undescribed species,"

62

4. Aleurodes vaporarium (?), Garman, Ann. Rept. Ky. Expr. Sta.r

. p. 37 (1890).

5. Aleurodes vaporarium (?), Garman, Agric. Science, vol. v, p.

264 (1S91).

6. Aleyrodes sp. (?), Riley, Insect Life, vol. v, p. 17 (1892).

7. Aleurodes sp. (?), Webster, Ann. Rept. Ohio Expr. Sta., p.

35 (1894).

8. Aleyrodes vaporariorum, Britton, 19th Rept. Conn. Expr. Sta.,

p. 203 (1896).

9. Aleyrodes sp. (?), Slingerland, Bui. 19, Cornell Expr. Sta., p.

^S-^S (i901)-

10. Aleyrodes vaporariorum (?), Britton, Bui. 140 Conn. Expr. Sta.,

P. 3, IO> !4, 17 (19°2)-

11. Aleyrodes packardi, Morrill, Can. Ent., vol. xxxv, p. 25-35

09°3)-

12. Aleyrodes vaporariorum (?), Britton, 2nd Rept. Conn. State

Ent., p. 149, 156, 160, 162 (1903).
it,. Aleyrodes packardi, Morrjll, Psyche, vol. x, p. 80,81,83 (I9°3)-

SUMMARY.

Of the sixty-five known American species of the genus Aleyrodes,
the two herein treated, the greenhouse Aleyrodes (A. vaporariorum)
and the strawberry Aleyrodes {A. packardi), together with the orange
Aleyrodes (A. citri), are the only ones that have thus far proved of
much economic importance. The orange Aleyrodes occurs only in
greenhouses on citrous plants in northern climates, and at present is
of no importance in this state.

The common names of the insects of this genus are : Aleyrodes,
white fly, mealy wing and snowy fly.

The family Aleyrodidas, to which the genus Aleyrodes belongs,
occupies a systematic position between the plant lice and the scale
insects, and appears to be more closely related to the latter.

63

The Greenhouse Aleyrodes.

The immature stages of this insect consist of the egg, three
larval stages or instars and a so-called pupal stage. Of these, only
the first larval instar is furnished with well developed legs and is able
to crawl. The pupal stage is characterized by wax rods of variable
length, which arise from the dorsal surface.

The egg hatches in from ten to twelve days, the larval and pupal
stages together last from twenty-five to thirty days, and the adults
may live, and as a rule probably do live, for several weeks.

Adult females are about three times as abundant as the adult
males. Unfertilized eggs will hatch, and, as far as known, develop
into adults of the male sex.

An adult female has been known to lay more than one hundred
and twenty-nine eggs in thirty-six days.

There seems to be no reason to doubt that the common green-
house Aleyrodes of North America is the same as that described by
Westwood in 1856 as Aleyrodes vaporariorum.

The insect is supposed to have come originally from Mexico or
Brazil, but it is at present very widely distributed in greenhouses in
Europe and North America. It has a wide range of food plants, in-
cluding many of such economic importance as cucumbers and
tomatoes.

The insect is generally believed, by those who have had experi-
ence with it, to be the most serious greenhouse pest known at the
present day.

Simple preventive measures may be all that it is necessary to
use in many cases to keep the greenhouse free from the insect.

Spraying tomato plants in greenhouses is to be avoided when
possible.

Hydrocyanic acid gas is the cheapest and most efficient remedy
for the Aleyrodes in greenhouses. At present, it is advisable not to
exceed the rate of .1 gram of Potassium cyanide per cubic foot of
space for three hours' exposure after sunset. Even in a loose house
this can be depended on to destroy all the insects except the eggs
and a few pupce. Twice this amount of Potassium cyanide per cubic
foot of space in a tight house, with the same conditions otherwise,
will not injure tomato plants in a reasonable state of vigor, while it

64

has destroyed red spiders on a cucumber plant without injury to the
plant. Three fumigations, using Potassium cyanide at the rate of
.01 gram per cubic foot of space, and with an interval of two weeks
between each fumigation, should practically rid the house of the
pest.

When undesirable to use hydrocyanic acid gas, a combination
of fumigation with Nicoticide and syringing with a solution of Bow-
ker's Tree Soap seems to be the best substitute.

The Strawberry Aleyrodes.

This insect has been known for many years, and has been re-
ported from various parts of the Eastern United States.

It was formerly believed to be identical with the greenhouse
Aleyrodes, but it is now known to be distinct. Differences between
the two species have been found in all stages except the egg.

This insect passes the winter in the egg stage on the under side
of strawberry leaves. The eggs hatch in the spring and the earliest
adults appear from the first to^the middle of May.

Many of the larvae are destroyed by frosts and by starvation,
due to the leaves to which they are attached and from which they
are unable to move, drying up before the insects reach maturity.

In some cases it proves a serious pest.

Preventive measures are necessary if the treatment is to be fol-
lowed by good results. These consist in using care in the location
of new fields in growing plants for propagation in nurseries, etc.

The simplest and cheapest remedies consist in treating the
plants in the nursery. The old wintered over leaves should be re-
moved not later than the first of May. As all the insects are con-
fined to these leaves up to this time the plants will be freed from the
insect in proportion to the thoroughness with which these are re-
moved. It is necessary to combine the preventives with this remedy.

If the insect becomes so abundant in the field as to require
treatment, spraying with kerosene emulsion or whale oil soap, using
an underspray nozzle, is recommended.

65

EXPLANATION OF PLATES.

The cuts of Plate I have been obtained through the kindness of Mr.
W. E. Britton, State Entomologist of Connecticut, who has used them in
Bulletin 140 of the Connecticut Agricultural Experiment Station, and in
his Second Report of the State Entomologist.

PLATE I.

*

Fig. A : Aleyrodes vaporariorum ; early stages on tobacco leaf. En-
larged about four times.

Fig. B : Aleyrodes vaporariorum ; adults and pupa skin on tobacco
leaf. Enlarged four times.

PLATE II.

Aleyrodes vaporariorum.

Figs. 1 and 2 : Eggs.

Fig. 3 : Dorsum of first instar.

Fig. 4 : Venter of first instar.

Fig. 5 : Vasiform orifice of first instar.

Fig. 6 : Dorsum of second instar.

Fig. 7 : Venter of second instar.

• Fig. 8 : Right antenna of second instar.

PLATE III.

Fig. 9
Fig. 10
Fig. 11
Fig. 12

Fig- 13
Fig. 14

Aleyrodes vaporariorum.

Right hind leg of first instar, from below.

Vasiform orifice of second instar.

Venter of third instar.

Dorsum of third instar.

Vasiform orifice of pupa.

Legs, antennas arnd mouth parts of pupa.

Fig. 15
Fig. 16
Fig. 17 :
Fig. 18:
Fig. 19:
Fig. 20
pupa case.

PLATE IV.
Aleyrodes vaporariorum.

Side view of pupa.

Dorsum of pupa.

Right hind tarsus of adult female.

Margin of wing of adult.

Abdomen of male from the side.

Abdomen of female from the side, soon after emergence from

66

PLATE V.

Aleyrodes vaporariorum.

Fig. 21 : Adult female from above.

Fig. 22 : Adult female from below.

Fig. 23 : Adult female from the side, abdomen distended with eggs,
ovipositor extended.

Fig. 24. Right antenna of adult female.

Fig. 25 : Genitalia of male from the side. .

Fig. 26 : Terminal segment of abdomen of adult male from above,
showing vasiform orifice and genitalia.

Fig. 27. Ovipositor of female from above.

PLATE VI.

Figs. 2S-31 : Diagrams illustrating variation in number and position
of wax rods of the inner submarginal and dorsal series of the pupa of Aley-
rodes vaporariorum.

Fig. 32 : Diagram illustrating main trunks of tracheal system of larval
Aleyrodes; drawn from a larva (third instar) of A. packardi seen from be-
,low. For explanation of lettering of figures see page 28.

Fig. 33 : Dorsum of the first instar of Aleyrodes packardi.

PLATE I.

A. Young of A. vaporariorum on tobacco leaf : enlarged about four times.

13. Adults and pupa skins of A. vaporariorum on tobacco leaf : enlarged four times.

PLATE II.

PLATE III.

hlnuK <kl

PLATE IV.

PLATE V.

PLATE VI.

28

29

33

lflffViMdU.

HATCH EXPERIMENT STATION

-OF THK-

MASSACHUSETTS

AGRICULTURAL COLLEGE.

TECHNICAL BULLETIN NO. 2.

THE GRAFT UNION.

OCTOBER, 1904.

The regular Bulletins of this Station 7vill be sent free to all news-
papers in the State and to such individuals interested in farming as
may request the same. Technical Bulletins are sent only to those persons
interested in the subject treated of in each case.

AMHERST, MASS.:

PRESS OF CARPENTER & MOREHOUSE
1904-

HATCH EXPERI1YEEBJT STATION

OF THE

Massachusetts Agricultural College,

AMHERST, MASS.

STATIONS! STAFF:

Henry H. Goodell, LL. D., Director.

William P. Brooks, Ph. D., Agriculturist.

George E. Stone, Ph. D., Botanist.

Charles A. Goessmann,Ph.D.,LL.U.,c7^///zj-/ (Fertilizers).

Joseph B. Lindsey, Ph. D., Chemist (Foods and Feeding).

Charles H. Fernald, Ph. D., Entomologist.

Frank A. Waugh, M. S., Horticulturist.

J. E. Ostrander, C. E., Meteorologist.

Henry T. Fernald, Ph. D., Associate Entomologist.

Frederick R. Church, B. Sc, Assistant Agriculturist.

Neil F. Monahan, B. Sc, Assistant Botanist.

Henri D. Haskins, B. Sc, First Assis't Chemist (Fertilizers).

Edward G. Proulx, B. Sc, Third Assis't Chemist (Fertilizers).

Edward B. Holland, M. S., First Chemist (Foods and Feeding).

Philip H. Smith, B. Sc, A ss't Chemist (Foods and Feeding).

Albert Parsons, B. Sc, Inspector (Foods and Feeding).

Joseph G. Cook, B. Sc, Assistant in Foods and Feeding.

George O. Greene, M. S., Assistant Horticulturist.

George W. Patch, Observer.

The co-operation and assistance of farmers, fruit-growers, horti-
culturists, and all interested, directly or indirectly, in agriculture,
are earnestly requested. The Bulletins will be sent free to all news-
l the State and to such individuals interested in farming as
f request the same. General bulletins, fertilizer analyses, analy-
of feed-stuffs, and annual reports are published. Kindly indi-
i^plication which of these are desired. Communications
may be addressed to the

Hatch Experiment Station, Amherst, Mass.

THE GRAFT UNION.

Some Observations on the Nature of the Union Between
Slock and Cion, Especially in Hard-
wood Grafting-.

BY F. A. WATCH.

Grafting is one of the most important processes in horticulture.
As soon as we penetrate its essential nature we find it also one of
the most intricate of subjects. The importance of the practice, and
the difficulties which stand in the way of a full understanding of it,
must make it always the subject of much speculation and study ; and
anyone who contributes, even a little, to make the knowledge of
graftage clearer, is doing some sort of a horticultural service. These
reflections must be the writer's excuse for presenting here certain
observations, the most of which seem to be so obvious and simple as
hardly to deserve statement in words. Nevertheless, in spite of
their obviousness, some of these simple facts have been commonly
overlooked, and in disregard of them, fallacious arguments have
been made and erroneous conclusions drawn.

We speak of graftage as the union of a cion with a stock. In the
vast majority of cases the prime object — we might say almost the
sole object — of graftage is to secure this union. The nature of this
union is generally understood to determine the whole success or fail-
ure of the graft. We say that the cion unites with the stock; and
we speak of good unions and poor unions. These phrases are al-
most the commonest in horticulture; yet their significance is gener-
ally unknown, the facts are wrongly guessed, and the whole nature of
the matter essentially misunderstood.

If this statement seems too sweeping, let us wait a moment until
some of the simpler facts can be stated. Doubtless there are persons
who have had the wisdom to see long ago all that is here set down ;
but many persons, even horticulturists of experience and ability,
have long overlooked the facts.

How do stock and cion unite in the growth of a graft ? The short
and easy answer is that the two grow together. The common idea
is that the two pieces grow fast to one another, just as two pieces of
bone grow together when they heal after a healthy boy breaks his
leg. There is a notion further that both cion and stock produce
new tissue, and that these new tissues commingle or coalesce in some
way ; but this speculation is even more vague than the primary one
which leads to it.

In herbacious grafting (where soft growing parts are used) the
union may possibly be somewhat of this sort ; but even here there is
usually no general commingling of the two members as has been
popularly imagined. The original cion and the original stock re-
main to the end of their existence very largely separate and dis-
tinct.

In ordinary graftage, where the cion is a cutting of hard, ripe, one-
year-old wood and the stock is a dormant stem or root, it is, on the
face of the proposition, impossible that the two should unite. Dis-
regarding for the moment the very thin (though very important)
cambium zone, the stock and cion are made up wholly of dead wood
and bark. With perfectly negligible exceptions the cells are all
dead — totally and forever dead. It is absolutely impossible for them
to grow or to unite with anything. One might as well talk of making
a lead pencil unite with a pen holder or a neckyoke with a single-
tree. The two may be glued together, waxed together, tied to-
gether ; but they can never unite.

It is said sometimes that the cambium layers of stock and cion
unite. This is quite a different matter, and though in the sense in
which it is usually presented, the statement furnishes no explanation
of the process, it leads us in the right direction. This thin sheet of
cambium, lying between the bark and the wood, is the only portion
of a tree stem which is really alive, the only portion which can grow,
nd so, necessarily, the part to which we must look for the beginning
of the graft union.

If, now, we examine a successful graft of several years' growth,
we shall at once discover two important facts: First, the one al-
ready propounded, that the old cion and stock have remained totally
distinct and separate, without the slightest promise of a union ; and,
second, that the new wood which has grown since the graft was made
is continuous. It is in complete, continuous, unbroken annual layers,
like those in an ungrafted stem. At any rate, the difference between
these annual layers as they grow over a graft junction and as they
grow on an ordinary stem is not an essential one. It may be set
aside for the present, but will be examined later.

Fig. i. Diagram of Cleft Graft 3 years old.

These statements may be better understood by reference to the
accompanying diagram, Fig. i. The black portions represent the
wood of the original cion and stock. The white portions represent
three annual layers of wood which have grown since the graft was
made. That this is a true statement of what really takes place may
be seen from the photographs, Figs. 2, 3, 4, which are taken
direct from the actual specimens, and are reproduced natural size.
In every case the old stock can be traced, more or less clearly, along
its junction with the cion, while the annual layers may be seen com-
pletely enveloping both. We have thus demonstrated the following
simple facts :

(1) Stock and cion do not unite like the two parts of a broken
bone. On the contrary they remain forever separate.

(2) Neither do the annual layers produced after grafting unite,
in the ordinary sense of that term. Each layer, under normal con
ditions, is complete and continuous.

.. - '• v-

5

"a

in

0

(3) In hardwood graftage, therefore, the "union" of stock and
cion is quite different, in its mechanical nature, from what our com-
mon speech would signify and from what it is commonly understood
to be.

These conclusions are so simple and obvious that we are slow to
admit their revolutionary character. Moreover, they are immediately
followed in the horticulturist's mind by a train of questions and
doubts which, though entirely subsidiary, do much to disguise the
force of the main facts. The horticulturist knows, for instance, that
the two varieties brought together by graftage retain their character-
istic qualities unmodified, or only slightly changed. If a pear cion
is grafted on a quince stock, then any bud above the junction will
bear pears, and any bud growing from below the junction will bear
quinces. There must be a division between the two kinds of wood.
What is the nature of this division ? How precise is it ? Is it ac-
companied by any mechanical strength or weakness ? These and
other similar questions press for answer.

It cannot be said that we are yet in a position to answer all these
questions. Possibly we can throw a little light on some of them.
But it must not be forgotten that all these considerations are en-
tirely above and beyond the more fundamental conclusions already
reached, and that the facts already developed are not to be in any
wise affected by the discussion of these questions which we now
take up.

THE SEPARATION OF CHARACTERS

If we make sections of a large number of grafts, such as are shown
in the accompanying photographs, we shall find that, in spite of the
longitudinal continuity of the annual layers, there is sometimes, at
right angles to them, a visible line of demarcation between the wood
grown from the cion and that grown from the stock. The two kinds
of tissue are sufficiently unlike that the difference can be noted with
the naked eye. Moreover, in some cases there is a distinct line
which seems to form a boundary between the two members.

n
P.

c

a)

u

O

c e

C o

S 5

c rt

One would naturally expect that a microscope having a sufficient
magnification to show the individual cells of which the tissues are
made up would quickly exhibit the manner of their conjunction.
However, after having studied a large number of such specimens
under the microscope, the writer is compelled to report the results
disappointing. In nearly all cases one can see less with the micro-
scope than with the naked eye.

Fit.. 4. Dyehouse Cherry on Miner Plum, showing continuous annual layers

of wood.

In figure 6, which is very greatly magnified, showing the cells
at something like 1000 times their natural diameter, the little knot of
scar tissue near the middle does accidentally show one point on the
line of junction ; but it will be seen at the same time that the vessels
and the parenchyma cells run uninterruptedly from end to end.

io

Fig. 5. Sections through Grafts of Plum.

With this photograph showing the microscopic anatomy of the
tissues before us, it is harder than ever to see where a certain tree
ceases to be cherry and begins to be plum. The cells are so in-
differently mixed that one is tempted to believe that the specific
physiological character may be blended at the same time. It
has long been the dream of gardeners to produce new kinds of plants
by the graft union of diverse cions and stocks. Do not our present
observations show the reasonableness of such endeavors?

Certainly not. No matter how closely the two kinds of cells may
lie against one another, their contents are never mingled in the pro-
duction of a new cell. Each new cell is produced by the division of
some older single cell, never by the fusion of two parent cells. Let
us suppose that in figure 7 the black cells are those of the quince
stock, while the white cells are those of the pear cion. As each

I I

' I ■ I

In,. 6. Section through a young graft, greatly magnified. The round mass of
scar tissue is merely accidental.

black cell divides quite independently of the white ones, it produces
only black offspring; and, similarly, the white cells dividing produce
only white offspring. Conversely stated, each cell is the offspring of
one parent only, not of two ; and it will therefore have the qualities
of one parent onlv.

The two kinds of tissue may commingle or lap in together some-
what along the line of junction, but this mixture is only mechanical,
not physiological.

The question of the mechanical strength of the graft union may
be postponed for a moment until we consider the morphology of the
union produced in budding.

I 2

Fig. 7. Diagram to illustrate separateness of character in stock and cion.

THE BUD UNION

Budding is simply a form of graftage in which the cion carries
only one bud, with little or no \vood. The form by far the most
commonly used is known as the shield bud. This is the only form
here considered, but other methods give strictly comparable results.

Figure 8 represents diagramatically the growth of a bud when
set upon a stock. The black portion of the figure represents the

Fig. S. Diagram of bud graft 3 years old.

wood of the stock and the shaded portion represents the bud or cion.
The white portions represent three annual layers of growth which

'3

have been put on since the bud was set. In this case, as with the
common graft, the layers of new growth are continuous, running from
top to bottom without any break at the plane of junction.

If a successful bud graft of several years' standing is cut in two
along its axis we shall find almost precisely the same conditions as
we have already found in the ordinary graft. Although the annual
growth layers are continuous, we may sometimes find a line of de-
marcation visible to the naked eye. When this is examined under
the microscope, however, the distinction usually disappears to a con-
siderable degree, and the cells are found to run together indistin-
guishably. There is therefore no difference, from our present point
of view, between a bud graft and a long-cion graft.

MECHANICAL STRENGTH OF THE JUNCTION

There has always been a good deal of debate amongst horticultur-
ists as to the mechanical strength of a graft. Some have said that
the graft always presents a point of weakness in the tree. Others

FlG. 9. Bud grafts 1 and 3 years old. showing continuous layers of young wood.

14

"have denied this, and have even said that the place where a graft has
healed successfully is the strongest point in the stem. The materials
which we have here in hand enable us to put this vexed matter in its
proper light.

It will be noticed that the region about some grafts is more or less
swollen by the deposition of an extra amount of wood tissue (due to
causes which we need not now discuss), and that the wood in this
region is very close grained, as may be seen on careful examination.
In some cases when grafts are cut open and dried the tissues crack
or check more quickly at other points than at the graft junction,
showing that the fibers are stronger in the latter region. It may also
be observed when a wind breaks off branches in an old orchard that
many, usually a majority, of the fractures occur, not where the grafts
have been made, but elsewhere on the same stems, a fact which
shows conclusively that the grafts are not points of weakness in those
particular cases.

On the other hand grafts do sometime break, even after they have
grown, in apparent health, for a number of years. When a branch
breaks at the point of graftage it is prima facia evidence that that was
the weakest point in that particular branch. Moreover there are
whole classes of facts which the horticulturist has assembled here.
Certain kinds of plants are known as a rule to make weak unions,
the Clairgeau pear on the quince and the Domestica plum on the
peach being examples. A certain nurseryman recently, on my order,
sent me ioo " poor unions." These were taken from his nursery
rows and were simply sorted out from his stock. In every case the
cion. had grown, but the union was bad, /. e., mechanically weak. If
the normal union of cion and stock is made by such complete and
continuous cylinders of annual growth, how can we account for the
unsuccessful unions ?

The answer is not altogether an easy one. We may approach it
by saying that, when the two members are unlike in nature and in
some way physiologically incompatible (whatever that may mean),
the wound does not heal readily, owing to some sort of irritation
which continues to be felt at this point. In nearly all cases this is
followed by the deposit of considerable quantities of loose meristem-
atic tissue, such as is always produced by the tree for the healing of a
wound. These loose and roundish cells fill the space where in a
successful graft the long parenchymous cells and ducts interlace.

The whole area of the union is therefore weaker.

The constant and excessive deposits of these soft roundish cells
is what causes the swellings so frequently seen at the point of union
in grafts.

In some cases these round cells form a thin wall of division be-
tween the tissues of the growing stock and cion. This was the case
with the bud graft illustrated in figure 9, a pear on a quince root,
which, after growing three years, broke off with a cleavage almost as
clean as the ball and socket of a hip joint. This articulation will be
seen filling the middle half of the area of fracture. The surround-
ing portions, broken more roughly, are the loose corky tissue which,
filling in between the wood and the bark, have formed the large
swelling at the point of junction.

v

Fig. 10. A defective and broken bud graft. Pear on Quince.

After a close study of a large number of these defective unions,
the writer has reached the opinion that they are almost always due
to the incompatibility of stock and cion, as referred to above. It is
a common notion among horticulturists that careless or ignorant
manipulation — the defective setting of bud or cion — in the grafting
will lead to these poor unions. There seems to be little ground for
this opinion. If the stock and cion are of varieties which are con-

genial, and if the graft or bud grows at all, the union will nearly
always be good. Poor manipulation will often cause the failure
of a large percentage to grow, but it seldom affects permanently the
strength of the union in those grafts which live at all.

GENERAL ESTIMATE

Two questions emerge from the foregoing discussion. The scien-
tist asks what effect these observations have on the philosophy of
graftage. The practical fruit grower asks how these facts bear on
his work of grafting trees. The writer does not believe it is neces-
sary for every fact to be theorized, nor yet that it should be given an
immediate practical application. The notions of graft unions hith-
erto entertained by most horticulturists have been, for the most part,
merely theories, more or less incorrect, and these incorrect theories
were the less excusable because the facts were so abundant and so
easily to be seen. It is hoped that the simple observations of this
bulletin may help to clarify the common views of graftage. No
direct application to the practice of graftage will be attempted here ;
but it may be said parenthetically that this department has in hand
several lines of experiments bearing on the practical aspects of
grafting and budding.

In conclusion we may re-state briefly the points developed:
(i) In hardwood graftage the cion and the stock never grow to-
gether.

(2) Neither do the new layers of wood grown from the cion and
stock respectively unite. Instead of this, under normal conditions,
they are produced in perfectly continuous layers.

(3) In the case of imperfect unions the continuity of the new
growth is more or less interrupted by the deposition of a certain
amount of loose scar tissue such as serves in the healing of wounds.

(4) These conditions make the graft union mechanically weak.
All degrees of mechanical strength may be observed in graft unions,
ranging from those (a large number) which are stronger than the ad-
jacent parts of the same stems down to such as are incapable of
even holding themselves in place.

(5) Incidentally it has been judged that weakness of unions re-
sults nearly always from physiological incompatibility of the cion
and stock, and very seldom from faulty manipulation in the original
setting of the graft.

Technical Bulletin No. 3. APRIL, 1907.

MASSACHUSETTS

AGRICULTURAL EXPERIMENT

STATION.

THE BLOSSOM END ROT OF TOMATOES

BY

ELIZABETH H. SMITH, M.S.

This Bulletin gives the results of investigations pertain-
ing to the cause of the so-called "Blossom End Rot" of
Tomatoes, together with a discussion of results obtained
by other investigators.

Requests for bulletins should be addressed to the

Agricultural Experiment Station,

Amherst, Mass.

MASSACHUSETTS AGRICULTURAL
EXPERIMENT STATION.

AMHERST, MASS.

COMMITTEE ON EXPERIMENT STATION :

Charles H. Preston, Chairman,

J. Lewis Ellsworth, *

William H. Bowker, Samuel C. Damon.

The President of the College, ex officio,
The Director of the Station, ex-officio.

ST^TIOUST STAFF:

CHARLEsA.GOESSMANN,PH.D.,LL.D.,/2r0«0rtfry Director and Chemist

(Fertilizers).

William P. Brooks, Ph. D.,
George E. Stone, Ph. D.,
Joseph B. Lindsey, Ph. D.,
Charles H. Fernald, Ph. D.,
Frank A. Waugh, M. S.,

J. E. OSTRANDER, C. E.,

Henry T. Fernald, Ph. D.,
James B. Paige, D. V. S.,
Henry J. Franklin, B. Sc, <.
Erwin S. Fulton, B. Sc,
Neil F. Monahan, B. Sc,

Henri D. Haskins, B. Sc,

*

E. Thorndike Ladd, B. Sc,
Edward B. Holland, M. S.,
Philip H. Smith, B. Sc,
Lewell S. Walker, B. Sc,
William K. Hepburn,
Roy F. Gaskill,
Carl S. Pomeroy, B. Sc,
Edwin F. Gaskill, B. Sc,
T. A. Barry,

Director and Agriculturist.
Botanist.

Chemist (Foods and Feeding).
Entomologist.
Horticulturist.
Meteorologist.
Associate Entomologist.
Veterinarian

Assistant in Entomology.
Assistant Agriculturist.
Assistant Botanist.
First Assist Chemist (Fertilizers).
Second Assist Chemist (Fertilizers).
Third AssisH Chemist (Fertilizers).
First Chemist (Foods and Feeding).
Ass't Chemist (Foods and Feeding).
AssH Chemist (Foods and Feeding).
Inspector (Foods and Feeding).
Assistant in Foods and Feeding.
Assistant Horticulturist.
Assistant Agriculturist
Observer.

* A vacancy exists at present.

Annual reports and bulletins on a variety of subjects are published.
These are sent free on request to all interested in agriculture. Part-
ies likely to find publications on special subjects only of interest will
please indicate these subjects. Correspondence or consultation on
all matters affecting any branch of our agriculture is welcomed.
Communications should be addressed to the

Agricultural Experiment Station,

Amherst, Mass.

DIVISION OF BOTANY.

The Blossom End Rot of Tomatoes/

ELIZABETH H. SMITH.

For several years tomatoes have been grown in one of the green-
houses of the Experiment Station, during which time what is known
as " Blossom End Rot " has been present in more or less abundance.
On account of the diverse results obtained by various investigators
of this trouble, an attempt at a diagnosis has been made by the
writer.

All the diseases of the tomato fruit now known may be summed
up under the terms " Black Rot," " Blossom End Rot," and " Fruit
Rot." Up to this time it has not been determined to what extent
these are identical.

REVIEW OF PREVIOUS WORK.

Although Blossom End Rot has been mentioned in various ex-
periment station bulletins, no careful investigation was made of the
disease before that of Galloway1 in 1888. He describes a " Black
Rot " of the tomato as follows : "The disease, as a rule, makes its
appearance at the apex or flower end of the fruit, when the latter is
from one-half to two-thirds grown. At first, a small blackish spot
is seen either around the remains of the style or on one side of it ;
this rapidly increases in size, but retains a more or less circular out-
line. As the disease progresses the tissues collapse quite regularly
on all sides, and the berry becomes much flattened. There is usually

* These investigations on tomato rots were accepted as a thesis for the degree of M. S.
The greater part of this work was done four years ago, and the manuscript in its present
form was completed in June, 1905. G. E. S.

2

a slightly raised narrow border surrounding the diseased parts, while
just outside this the cuticle retains its normal healthy color, but ap-
pears slightly wrinkled, owing to the collapsed condition of the tissues
beneath. Sections through a rotten tomato at this stage show that
the black discolorations extend deeply into the tissues, the depth
depending somewhat on the size of the spot. As the malady pro-
gresses, the diseased parts become hard and leathery, the surface
assumes a greenish black velvety appearance, and finally the entire
fruit becomes dried and shriveled."

Galloway identified two species of fungi in connection with the
spot, viz.: Macrosporium to?nato, Cook, and Fusarium so/am', Mart.,
both of which he describes in detail. He finds the spores of the former
exceedingly variable, the hyphae at times long and nodulous and bearing
spores which cannot be distinguished from those of the genus Clado-
sporium. Galloway finds that the Macrosporium grows mostly in a
felty mass near the surface, while the Fusarium shows a much greater
tendency to penetrate the sound tissues.

The results of Galloway's infection experiments are summed up as
follows: (i) Neither the Fusarium nor the Macrosporium has the
power of penetrating the sound cuticle of the tomato. (2) The
Macrosporium spores, when brought in contact with the exposed
tissue of either green or ripe fruit, produce the rot in a very short
time. (3) The Fusarium will grow only in fully ripe tissues or
tissues which have been partly disorganized through other agents.

From the publication of Galloway to that of Jones and Grout3 in
1897, tomato rot has been often mentioned in experiment station
bulletins, and Macrosporium tomato, or by error M. so/am, has been
given as the cause. A study of the so-called Macrosporium sotani of
potatoes by Jones and Grout resulted in the identification of two
species, one an active parasite and one saprophytic. Both species
were proved to be Alternaria. " When the study of the tomato rot
was taken up (by the same workers) it was found that the fungus
which causes the black patches on the rotting fruit was an Alternaria,
not distinguishable from the one on onion and potato. It was
further proved that this fungus did not cause the rot, for green
tomatoes inoculated with spores from a pure culture from the tomato
remained fourteen days in a moist chamber until they ripened, with-

out showing signs of the rot, and cultures made from tomatoes just
beginning to rot would not yield the Alternaria."

In an article on Blossom End Rot in the English " Journal of
the Board of Agriculture," by Charles Whitehead2, the cause of the
disease is attributed to Cladosporium lycopersici, " This rot," it is
stated, " must not be confounded with the other tomato affection
known as ' Black Rot,' due to another fungus termed Macrosporium
tomato, Cook. This is not confined to the lower part or style end of
the fruit. The color of the rot is altogether much darker than the
decay consequent on the attack of Cladosporium lycopersici.'''' The
same worker noticed another fungus in tomatoes infected with
Cladosporium lycopersici, having a mycelium more delicate and slen-
der than that of Cladosporium, and quite colorless. This permeated

Fig. i. Section of the natural spot showing dead
cells, masses of starch and mycelium of Fusarium.

the tissues of the fruit and lived upon it, penetrating further into the
healthy tissues than the Cladosporium, though the Cladosporium
was " evidently the originator and main cause of the mischief."

The same writer states that Dr. Plowwright in 1881, presumably
in the same journal, reported both a Cladosporium and a Macro-
sporium in connection with the disease, the Macrosporium being
identical with Macrosporium tomato, Cook, of Galloway, previously
quoted.

In 1900, F. S. Earle\ of the Alabama station, published the results
of work on the " Black Rot or Blossom-end Rot," as follows :

" When tomatoes are attacked by this disease in the field, the first
stage to be noted is the appearance of a small, irregular, watery

area, usually, though by no means always, surrounding the remains
of the pistil. On making a cross section this watery condition is
found to be confined to the portion immediately under the skin. It
usually does not involve the tissues to any great depth, even after it
has extended so as to cover a considerable surface area. Growth of
the fruit over the infected area stops, so that after a few days the
spot seems somewhat sunken. If the tomato is nearly ripe, maturity
will be hastened and the watery spot may dry down so as to look as
if the fruit had been slightly seared with a hot iron. The greater
number of infections take place when the fruit is about an inch in
diameter. The disease may invade the entire surface, causing the fruit
to fall, or the premature ripening of the lower portion may arrest it,
when the partially dried diseased portion often becomes blackened
by a velvety growth of the Alternaria. In the early morning drops
of sticky exudation were observed on the spots of half-grown rotten
fruit. These were found to be swarming with bacteria, which were
found abundantly within the diseased tissues. Sound green toma-
toes under a bell-jar were inoculated with a pure culture prepared
from the exudate. In all cases they showed signs of rot within
twenty-four hours. When agar containing the germs was smeared
on the surface of sound tomatoes, no rotting took place even after a
number of days. The disease cannot be contracted through the
flowers, as is the case in pear blight. The stigmas of many open
flowers were smeared with cultures of the germ without inducing a
single case. In no case were inoculations successful where the fruit
was less than one centimeter in diameter. It grows on ripe toma-
toes, but less readily than on green ones. It seems to be strictly
aerobic."

From the abundance of thrips on the fruit, he concluded that in-
sects may be the means of dispersal of the germ.

William Stuart5, of the Indiana station, reported a tomato disease
in 1900, which, he says, " resembles in many respects that caused by
Macrosporium solani." He attributes the rot to a motile bacillus
found in the tissue. He is doubtful, however, of having isolated it,
as inoculated fruit developed a watery rot with an offensive odor,
which soon decomposed the fruit completely. The natural " spot "
was perfectly clean, and is described as a watery discoloration, which
afterward dried down.

FUSARIUM ROT.

Work on the tomato fruit rot was begun by the writer in the spring
of 1902. A diseased condition of the fruit was observed at once on
the maturing of the crop, the appearance of which agreed in a gen-
eral way with that of Galloway1 and Whitehead2, previously
quoted. The spot was always around or at one side of, and contin-
gent with the style, slightly sunken, with regular boundary, and
clearly defined by a slightly raised ridge, beyond which the tomato
was perfectly healthy. The tissue was dry and leathery, a light
greyish brown, showing more or less of the shriveled vascular bun-
dles beneath. On the larger spots rings of a darker color were
sometimes barely visible near the boundary. Small, dark spots were
sometimes present on the rotten area. The longitudinal section
through a spot showed the interior boundary to be as clearly defined
as the outer. The depth of the rotten area was about equal to the
surface radius in the most extreme cases, and often extended not
more than half that distance.

A microscopic examination of the spot showed nothing which
could be distinguished as a fungus until a careful search had been
made. The tissue was dry and blackened, the cells disintegrated.
Finally a small tuft of delicate, colorless, mycelial threads was ob-
served along the inner border, and by careful searching this charac-
teristic form was found in every spot, most easily at the interior
margin of the rot. The tufts protrude from intercellular spaces be-
tween the blackened walls into the still healthy tissue. (See Fig. 2.)
Tufts were also found in the interior of the rot, but in most cases
were shriveled beyond recognition.

A study of the spots on ripe fruit showed the same characteristics
as on the green fruit, but it was found that some of the large starch
grains so abundant in the green fruit had been retained in the cells,
while the healthy tissue, as usual, gave no reaction with iodine for starch.

Particular attention was paid throughout to the detection of any
bacteria which might be connected with the disease. The spot, when
first opened, was in all cases perfectly dry and apparently free from
bacteria. After remaining in a moist chamber for a day or two, a
dense growth of Fusarium developed on the cut surface, often accom-
panied by putrefactive bacteria. Repeated attempts to inocculate
with these bacteria failed, and several attempts to obtain cultures
from the fresh spot were unsuccessful.

8

An examination of the fungus showed it to be a species of Fusa-
rium. Spores were produced abundantly and proved to be the
characteristic Fusarium type — club-shaped in the young stage ; more
or less elliptical when mature ; usually four-celled, varying to six;
size varies from 32 — 38.4 fi in length, and the conidiophores are
short and thick. (See Fig. 3.) In none of the rot examined was
there the slightest trace of any other fungus.

Pure cultures of the Fusarium were obtained on stewed tomatoes,
and these were used to inoculate green fruit. Inoculations were
tried both around the style without penetrating the epidermis, and
through a wound. The Fusarium placed around the style did not
take effect, except in one case in which a small, quite characteristic
spot was obtained in this way on the green fruit.

Fig. 2. Mycelium from infected cells pene- Fig. 3. Conidia of Fusarium.

trating healthy tissue.

Inoculations through punctures in the epidermis grew with varying
vigor. When deep punctures were made the mycelium grew rapidly
until nearly the whole fruit became involved. The tissues affected
were softer, more collapsed and lighter in color than in the case of
natural infection, as was to be expected from the vigor of the growth.
Here, also, the tissues were perfectly dry. The surface of the rot
became covered with white concentric rings. The mycelium in this
case permeated the tissues in all directions, especially in strands be-
tween the cells. Spores were produced throughout, especially in
the intercellular spaces, (See Fig. 4.) The concentric rings proved
to be spore clusters developed in dense, conical tufts on the surface.
(See Fig. 5.) Macroconidia have not been observed in connection

with this species. By inoculations of varying depths, grrdations
in development between the natural rot and that just described were
obtained, showing that the fungus is aerobic.

Fig. 4. Mycelium of Fusarium in?deeply inoculated tomato,
showing conidia.

Spotted tomatoes were left on the vines
for weeks after ripening, but no further growth
of the spot took place. The first indication
of decay was the usual watery softening of
the tissue, leaving the spot as dry as pre-
viously. As before stated, the rot appeared
with the first fruits. It continued for about
eight weeks, and then disappeared entirely.
It was then about the middle of May, and the
plants were on the decline. Galloway states
that, " however long continued, the rot always
begins on the first fruit." This is the only
previous observation made as to duration.

Tf^e smallest infected fruit found by the
writer measures three-fourths of an inch in
diameter, and inoculation experiments with
fruit under this size have been unsuccessful.
These results agree in a general way with
those of Galloway and Whitehead. In the
former case the fruit was one-third to one-

Fig. 5. Filaments from a
conidial tuft from surface
of inoculated tomato.

IO

half grown, in the latter the size of a horse-chestnut. Ten inocula-
tions of Whitehead were not successful where the fruit was less than
one centimeter in diameter.

CULTURAL RELATIONS OF THE FUSARIUM.

Following as far as practical the method of Prof. R. E. Smith6 in
his work on Botrytis, cultures of the Fusarium were made on media
of such vegetable substances as it was likely to meet with during
its growth on the fruit. The mineral stock solution used was as
follows :

Potassium nitrate, 5.0 grams

Magnesium sulphate, 2.5 grams

Potassium phosphate, 2.5 grams

Calcium chloride, 1.0 gram

Water, 1000 c.c.

More nitrogen was supplied in the form of peptone. Each of the
following substances was added separately to 200 c.c. of the stock
solution in the proportion indicated :

Malic acid, .5$, 3$, iofc ; oxalic acid, .5$ ; tartaric acid, 2$ ; tannic
acid, 2% ; cellulose, small quantity ; corn starch, 2^ ; glucose, 2$ ;
cane sugar, 2^. As usual, duplicate flasks were prepared in each
case which were left uninoculated for comparison.

Malic acid. This was the first acid tried, and as some species of
Fusarium are known to grow best on an acid medium, a variety of
strengths were tested. The acid of the tomato has been stated to
be approximately .5$ malic. Contrary to expectation, there was no
growth in the two stronger solutions, very slight and long deferred
growth in the .5^.

Oxalic acid. Growth fairly good. Slow in starting.

Tartaric add. No growth for a week or more, then very slight
growth.

Tannic acid. A little slow in starting, but finally a vigorous
growth. After the surface was covered, a darkening of the solution
took place from the top downward, turning from a light yellow to a
deep wine color.

Glucose. Growth vigorous from the start.

1 1

Cane sugar. (Commercial.) Growth similar to that in glucose.

Cellulose. (Pure cellulose filter paper was used in sufficient amount
to soak up the liquid.) Growth vigorous at the start, though on the
whole not quite as vigorous as on glucose and cane sugar.

Starch. (Commercial corn starch.) The best growth from the
beginning. The undissolved starch in the bottom of the flask was
observed to decrease materially, while the mycelium seemed to grow
down into it in a compact mass. It is evident from this result that
starch is easily available to the Fusarium as food, probably, it was
thought, by a rapid conversion of the starch into glucose by the
means of diastase excreted by the fungus.

To investigate this matter more fully, a concentrated solution of
starch was made by taking a small amount of starch and allowing it
to stand in water for a day. A mycelium extract was then prepared
by grinding a considerable amount of the mycelium in a mortar with
a small amount of clean sand. Water was then added, the whole
poured off the sand and allowed to stand for a day or more. The
liquid was then decanted and used. Five cubic centimeters of the ex-
tract was added to twenty cubic centimeters of the starch. In one flask
the extract was boiled, and in another unboiled. After a number of
days two cubic centimeters of a very dilute solution of iodine was
added to each flask. In each case a deep blue color was obtained, if
anything lighter in the unboiled flask. As it was evident from cultures
that starch is readily utilized by the fungus, some of the starch solution
was inoculated with the Fusarium. After a week's growth, the iodine
test was again tried. In this case, there was no reaction, showing
that all the starch had been converted and utilized by the fungus.

One of the reasons for the preceding cultural experiments was to
ascertain, if possible, why the fruit became infected at an early stage
only, while the rot failed in every case to increase after the tomatoes
were more than about i}( inches in diameter.

The result of the starch experiment would seem to throw some
light on the subject. If Fusarium solani grows particularly well on
starch, it would naturally develop while the starch is present in
greatest abundance. However, this does not explain why the rot
never appears on fruits nearly or fully grown, but which have not
lost their chlorophyll. The malic acid experiment suggested a pos-
sible clew to this. If, as shown by the experiment, the acid, when

I 2

present in quantities as great as t>% checked the growth of the
Fusarium it was thought that a determination of the relative acid in
green fruits would be valuable. Fruits of three sizes were selected
and tested, with results as below :

ACIDITY OF JUICE IN TERMS OF MALIC ACID.

Fruit Diameter. Amount of Acid.*
24 inch, 0.42$)

i}( inch, °-Sl$

2V2 inch, <

The minimum size at which the fruit becomes infected is three-
fourths of an inch in diameter, while 2^ inches was the maximum
in these fruits before changing color. The acid, then, increases in
proportion as the fruit grows. We may certainly consider this as a
significant fact in explaining the arrest of the fungus.

To ascertain the effect of any possible excretion of the mycelium
upon green tomato tissue, microscopic sections of the tomato were
treated with the mycelium extract. No immediate change took place
in the tissues. After about six hours, however, the cells were killed,
the protoplasm becoming slightly yellowed, shrunken and curled up
at the edges, and later the walls became blackened.

The Fusarium was grown in Petri dishes on the following cooked
vegetables : Potato, carrot, beet, parsnip ; and corn in the form of
meal. The growth was quite uniform in all, though perhaps a little
more vigorous on potato than on carrot and corn, while that on beet
and parsnip was slightly retarded. In all cases the tissues were
blackened after becoming permeated by the fungus. The fungus is
doubtless the Fusarium so/am', Mart, of Galloway.

OBSERVATIONS ON OUTDOOR TOMATO CROPS.

During the summers of 1902 and 1903 some work was done by
the writer on outdoor tomato crops. Here there was no difficulty in
finding all of the fungi recorded by writers previously quoted. In
many cases the spot was entirely covered almost from the first with
a dense growth of Macrosporium so/aui, but almost as often the

* These results were furnished by the department of Foods and Feeding of the Experi-
ment Station.

13

black growth of Macrosporium was confined to a portion of the
spot, the remainder being covered with a lighter green growth, which
proved to be Cladosporium. Chains of apical spores characteristic
of Alternaria were also found. In all three cases the mycelium grew
in a felty mass near the surface, agreeing in every respect with that
described by Galloway and others previously quoted. The average
size of the outdoor spot was perhaps larger than that in the green-
house. Cultures were made of these fungi and from them green fruit
was inoculated, both on the vine and under a bell-jar, but all inocu-
lations failed to infect.

BACTERIAL ROT.

In the late spring of 1904, another crop of greenhouse tomatoes-
was set out for the same work. The spot was present in abundance
from the first and lasted through August, viz., until the plants were
on the decline, as before. Out of more than a hundred diseased
tomatoes which were cut up and placed in a moist chamber, only ten
showed signs of Fusarium growth, and none of these were covered
with the vigorous felty mycelium which always developed in 1902.
In only one case was a pure growth obtained, the Fusarium being
slow to appear and accompanied by various molds, as well as bac-
teria. Again a pure culture of the Fusarium was obtained and
twelve inoculations made, all of which produced characteristic spots.
The constant and early appearance of the bacteria led to a trial of
these for inoculation.

CULTURAL CHARACTERISTICS.

The various Petri dish cultures taken from the rotting tomatoes
were inoculated by piercing the tissue just beneath the skin with a
flamed needle, after cutting through the spot with a sterilized knife.
Out of two hundred or more cultures made during the season one
characteristic organism was isolated. The deep and surface
colonies showed different forms as described below :

Description of the colonies in agar plate cultures :
Deep colonies — Round — fusiform, often elliptical, entire, porcelaneous
changing to butyrous, refraction strong, homogeneous,
coarsely granular, size .17 — .05 millimeters after
twenty-four hours.

*4

■Surface colonies — Round, convex, refraction weak, grumose, entire,
butyrous, with nucleus.

The growth of the organism in various media is as follows :

Agar-Agar — Comparatively slight growth, white.

Gelatin — Liquification, crateriform changing to stratiform in three

days.
Litmus gelatin — A band of yellow through the center of medium after

four days; after nine days, white with bluish ring at

top.
Litmus lactose agar — Heavy cream colored growth ; medium blue

after three days.
Milk — No coagulation.
Litmus milk — A slightly alkaline reaction.
Potato — In three days a thick, firm, tawny growth ; slight growth

after.
Slices of green tomato — Dense growth of a tawny color, darkening

with age.

From pure cultures of the above organism, sixty-seven inoculations
were made during August, and many which were unrecorded before
that time. Out of the sixty-seven, fifty inoculations produced a well
developed and characteristic spot within a few days. (See Fig. 6.)
Of those made in July a larger percentage was successful, and there
was also more rot by natural infection. From the spots artificially
produced, the organism was isolated repeatedly.

The position and general appearance of the spot during this
season were identical with that of 1902. The boundaries were clear
cut, and the surface, until late in the season, free from fungi. In
several cases, however, late in August, the fungi occurred as on the
outdoor crop, and again unsuccessful attempts were made to inocu-
late with these. There were slight differences to be noticed in the
color, consistency and markings of the rot in the two crops under
consideration. In 1904, the average spot was darker, there were
practically no concentric rings or veins visible on the surface, and
the tissue was often tender rather than leathery.

The successful inoculations with two organisms led to the idea of
trying others not taken from the spot. As work on soil bacteria was
going on in the laboratory at this time, three of these cultures were

J5

used for inoculation. Six inoculations were made from each culture.
The inoculations made with two of these were unsuccessful, but
those in which the third was used produced a rot which was not to
be distinguished from that by natural infection. After this, several
clean punctures were made in tomatoes on the vine by means of a
flamed needle. No rot set in for a week or more, after which a very
characteristic spot developed in two of the tomatoes which were
close to the ground. It cannot be definitely inferred in either of
these experiments that the rot was caused by an organism other than
the one studied, as no attempt was made to isolate an organism
either in the case of the soil bacteria or the clean punctures.

Fig. 6. Photograph of three tomatoes : One dis-
eased by natural infection, one by inoculation with
bacteria and one uninfected.

DISCUSSION OF RESULTS.

From the investigations recorded, we draw the following con-
clusions :

First, That Fusarium so/an/, Mart., is the initial and in all proba-
bility the only active parasitic fungus connected with the fruit rot of
tomato. The reasons for this may be summed up as follows :

i. Both Galloway1 and Whitehead2 determined a Fusarium in
connection with the disease. In both cases it is described as pene-

i6

trating the healthy inner tissues, while the accompanying fungus
remained more or less on the surface.

2. Galloway's record of the inoculation of green tomatoes with
Macrosporium is considered doubtful by Jones and Grout3, who
state conclusively that this fungus, called by them Alternaria, will
not grow on green tomatoes.

3. Whitehead2 states that Cladosporium is "evidently " the cause
of the disease, without apparently having attempted a proof by in-
oculation experiments.

4. Having found a Macrosporium, a Cladosporium and an Alter-
naria in connection with the rot of tomato, the writer has thoroughly
tested all by inoculation experiments, but was unable to induce any
of these fungi to grow on green tomatoes.

5. The writer, in a crop of tomatoes free from surface fungi, has
found Fusarium in the rotted tissue, which, when used for inoculating
green fruit, in every case produced a more or less natural spot. No
previous record of inoculation experiments with Fusarium has come
to the notice of the writer.

6. Starch is the best medium for the growth of Fusarium found
by the writer, it occurring in great abundance in green tomatoes.

Second, That the tissue of green tomato is a medium easily utilized
by fungi as food. That a number of fungi are instrumental
in producing rots which are often hardly to be distinguished from
each other. This conclusion has been reached from a study of the
following facts :

1. Both Earle4 and Stuart5 report tomato rots which they find to
be caused by bacteria. In the former case surface Macrosporium is
reported, but in the latter no fungus of any kind was detected. The
descriptions of the two rots are very similar, though one is appar-
ently more " watery " than the other. In both cases the organism is
partially described, but the work is insufficient for comparison.

2. The writer has studied a bacterial rot of tomato in which the
organism has been isolated and infection induced by repeated in-
oculation experiments.

3. Inoculations were made with three bacterial organisms taken
at random by the writer, from one set of which a rot developed which

'7
was not to be distinguished from the characteristic " Blossom End

'6

Rot."

Third, That the organisms causing the tomato rot are present in
the air and come into contact with the inner tissues of the tomato,
probably through cracks in the epidermis which occur about the
pistil. This seems to be the prevailing opinion in regard to the
infection of the fruit. Galloway advises a restriction of barnyard
manure as fertilizer, as he considers that it tends to increase the
cracking. Certainly the epidermis has become more tender with the
cultivation of the plant, and that certain varieties show this charac-
teristic more than others has been observed by the writer.

A number of infection experiments have been tried by spraying
the fruit with the bacterial organism grown in bouillon. Two or
three spots occurred which may have been caused by this treatment,
but the numbers were not sufficient to give conclusive results.

Fourth, That the quality of the fruit may affect the size, appear-
ance and position of the spot. This is demonstrated very nicely
in a greenhouse crop now under observation. On normal fruit the
rot is of slight depth, with collapsed tissue and clear boundary, as
found in 1902. However, the fruit in this crop is very largely lobed
and corrugated, and the rot in these cases extends almost to the
calyx before the surface begins to collapse, or even to change color.
On cutting open the tomato it is found that the carpels are imper-
fectly united at the pistil. In some cases quite a considerable open-
ing has been detected, through which the organism has reached the
watery and air exposed pulp. The pulp is often entirely dried and
blackened before the pericarp begins to be affected.

This may account for some differences in the records of previous
writers. The " seared " spot of Earle4 has been described by him
as occurring only on nearly matured fruits, to the ripening of which
he ascribes the arrest of the rot. (The spot referred to extends
through the epidermis only, turning it a light brown color.) We
have had many cases of the kind, which occurred quite as often on
the small as on the large fruit, but always on smooth, normal speci-
mens with no crack visible to the naked eye. Cases of entire rotting
have been found by us only on irregular fruit, as was also often
found by Earle and Galloway.

i8

In several cases during this season a small spot has appeared on
very abnormal tomatoes at some distance from the style. This posi-
tion has been mentioned in previous literature, but had not before
been observed by us. In these cases the rot was found to have
started from the pistil and in the pulp of one or more carpels, whence
it had spread to the pericarp at the point where the spot appeared.
Similar spots in this position were found to have been caused by
bruising of the fruit.

SUMMARY OF RESULTS.

i. Fusarium solani, Mart., may occur in connection with tomato
rot, unaccompanied by any other fungus.

2. A tomato rot with the same characteristics may be produced
artificially by inoculation with the Fusarium.

3. The same may be produced by ..placing Fusarium about the
style. (This occurred in only one case.)

4. Fusarium grows readily on most of the substances found in
green tomatoes, but less readily on sugar than starch, and not at all
in malic acid above 3$.

5. Starch is converted by the fungus in direct absorption rather
than by excreted diastase.

6. Fusarium solani will grow readily on all of our root vegetables,
but most vigorously on potato.

7. Mycelium extract from Fusarium has the power of killing pro-
toplasm and blackening the cells of green tomatoes.

8. Tomato rot may occur free from fungous growth, but contain-
ing a characteristic bacterial organism.

9. The same may be reproduced both by inoculation and by
spraying with the organism in bouillon culture. (The latter was
successful in two cases only.)

10. The organism isolated is an aerobic, spore-producing, mono-
trichic, oval form, producing a yellowish color on most media.

In the first part of the work here recorded much assistance was
received from Prof. R. E. Smith, now of the University of Califor-
nia. The work on bacterial rot was done under the direction of
Prof. G. E. Stone, by whom many of the experiments have been dup-
licated.

l9

LITERATURE CITED.

i. B. T. Galloway. Notes on Black Rot of the Tomato.

Rept. U. S. Dept. Agr., 1888, p. 339.

2. Charles Whitehead. Blossom End Rot of Tomato.

The Journ. of the Board Agr., Vol. Ill,
London, 1896, p. 154.

3. L. R. Jones. Potato Diseases and Remedies.

Vt. Exp. Sta., 10th Ann. Rept., 1896-7

PP- 5°> 51-

4. F. S. Earle. Black Rot or Blossom End Rot.

Ala. Exp. Sta., Bull. 108, 1900, p. 19.

5. William Stuart. A Bacterial Disease of Tomatoes.

Ind. Exp. Sta., Rept. 13, 1900, p. 13.

6. R. E. Smith. The Parasitism of Botrytis cinerea.

Bot. Gaz., June, 1902, p. 421.

lu-'v