Miscellaneous studies in agriculture and biology

SEMICENTENNIAL PUBLICATIONS

OF THE

UNIVERSITY OF CALIFORNIA

1868-1918

MISCELLANEOUS STUDIES

IN

AGRICULTURE

AND

BIOLOGY

UNIVERSITY OF CALIFORNIA PRESS

BERKELEY

1919

Q

in

CONTENTS

PAGES
A Synopsis of the Aphididae of California, by Albert

F. Swain 1-221

Mutation in Matthiola, by Howard B. Frost 223-333

Ocean Temperatures, their Relation to Solar Radiation

and Oceanic Circulation, by George F. McEwen 335-421

Changes in the Chemical Composition of Grapes during
Ripening, by F. T. Bioletti, W. V. Cruess, and H.
Davi . .. 423-450

A SYNOPSIS

OF THE

APHIDIDAE OF CALIFORNIA

BY

ALBERT F. SWAIN

[University of California Publications in Entomology, Vol. 3, No. 1, pp. 1-221, pis. 1-17]

A SYNOPSIS

OF THE

APHIDIDAE OF CALIFORNIA

BY

ALBERT F. SWAIN

CONTENTS

PAGE

Introduction 2

Classification 4

External anatomy 5

Biology 8

Economic considerations 9

Synopsis 10

Family Aphididae 10

Subfamily Aphidinae 11

Group Callipterina 12

Tribe Phyllaphidini 12

Tribe Callipterini 16

Tribe Chaitophorini 32

Group Lachnina 39

Tribe Pterocommini 39

Tribe Lachnini 43

Group Aphidina 52

Tribe Macrosiphini 52

Tribe Aphidini 87

Subfamily Pemphiginae 138

Group Hormaphidina _ 139

Group Pemphigina 140

Group Schizoneurina 147

Group Vacunina 150

Subfamily Phylloxerinae 151

Group Chermisina 151

Group Phylloxerina „ 152

2 MISCELLANEOUS STUDIES

PAGE

Appendix 1. Keys to the genera and tribes of Aphididae; a translation from

P. Van der Goot _ 154

Appendix 2. Host plant list _ 159

Addenda - — . 178

Explanation of plates _ 180

Index to genera and species 215

INTRODUCTION

In recent years considerable attention has been paid to the
Aphididae in the United States, and in Europe as well, and a large
amount of literature is the result. In California, W. T. Clarke was the
first to make any systematic studies of these insects and his paper,
published in 1903, embodies the results of these studies. He listed
forty-three species, ten of which were described as new. Two or
three of his new species are known at present, but the remainder are
unknown. Unfortunately his collection was destroyed in the earth-
quake of April, 1906, so now it is practically impossible to determine
his new species with any degree of accuracy. Following this, there
was a period of six years in which there were no publications con-
cerning the Aphididae of California, except some economic bulletins
from the Experiment Station. In 1909, both E. 0. Essig and W. M.
Davidson published the results of their earlier studies. Since then
both have added papers occasionally. During 1912 and shortly before,
Harold Morrison made an extensive study of the species in the vicinity
of Stanford University. He has kindly placed a report of his studies
in the author's hands, with permission to publish the records in this
paper. The author has been studying the Californian species con-
tinuously since 1914.

At present there are about one hundred and eighty species known
to occur in California. This number will undoubtedly be greatly
increased as further studies are made, since to date only a compara-
tively small part of the state has been covered by collectors. Very
extensive collections have been made in Ventura County and in the
vicinity of Pomona College, Los Angeles County, by Essig. The San
Francisco Bay region, particularly in the vicinity of the University
of California and Stanford University, has been carefully surveyed
for aphids, collections having been made by Clarke, Davidson, Essig,
Morrison, Ferris, and the author. Davidson, Clarke, and Essig have
made a few observations in the Sacramento Valley, particularly in
Placer and Sacramento counties. The author made a number of

A SYNOPSIS OF THE APHIDIDAE 3

observations in the vicinity of Fresno during May and June, 1915, and
more or less extensive observations and collections during 1916 and
1917 in San Diego, Riverside, Orange, Los Angeles, and San Bernar-
dino counties. In addition to these, reports come to the College of
Agriculture occasionally from the State Insectary and the various
county horticultural commissioners. A summary of the above state-
ments shows that extensive collecting has been done only in the
territory adjacent to San Francisco Bay, and throughout southern
California. The whole northern half of the state, the great interior
valleys of the Sacramento and San Joaquin rivers, and the desert
sections of the southeastern part of the state are as yet unexplored.
Undoubtedly many interesting species will be found in these parts.

The author wishes to express his appreciation of the aid rendered
by various people during the past three years of study. To Harold
Morrison of the Federal Board of Horticulture is due especial thanks
for his assistance during the early part of the author's study, for his
collection notes, and for the use of his extensive collections of Stanford
University vicinity and Indiana; Jo E. 0. Essig of the University of
California for his continuous advice and assistance, for the use of
his large collections of Californian species, and for the reading of
this manuscript; to W. M. Davidson of the Bureau of Entomology,
U. S. Department of Agriculture, for his many notes and deter-
minations and for the use of his collection; to A. C. Baker, J. J.
Davis, C. P. Gillette, A. S. Maxson, E. M. Patch, and H. F. Wilson
for their many determinations and suggestions ; to R. W. Doane of
Stanford University for the permission to work over his collection
of Utah aphids and for permitting his students to use the keys included
in this paper, thereby finding the weak points in the keys ; and finally
to G. F. Ferris of Stanford University for collections and advice.

In this paper the author has brought together all the present
records of California Aphididae. He has included keys for the
determination of the subfamilies, groups, genera, and species, together
with such illustrations as are necessary for an understanding of the
keys. The discussion of each species includes a bibliography of thf
California literature (exclusive of the merely economic and popular),
together with a citation of the original description and the best
available description, a list of host plants and localities, and a dis-
cussion of the synonomy, life history, and habits so far as they are
known. The descriptions of certain species are not readily accessible
and of others not at all adequate. Such species have been redescribed

4 MISCELLANEOUS STUDIES

by the author in so far as it was possible to obtain specimens. Inci-
dentally it may be stated that the author has personally collected by
far the larger number of the species recorded in this paper. In other
cases the fact is noted. A host plant index (appendix 2) is also
included.

The system of classification followed is the one most generally
accepted by American aphidologists at the present time. The keys
to the species have been formulated by the author, those to the genera
and higher groups have to a large extent been adapted from other
workers, particularly Wilson and Essig (Aphidinae), Borner (Phyl-
loxerinae), and Tullgren (Pemphiginae). The papers of Baker,
Clarke, Davidson, Davis, Essig, Gillette, Oestlund, Patch, Pergande,
Williams, Wilson, and other American aphidologists have been found
invaluable. Of the works of the European aphidologists, those of
Borner, Buckton, Del Guercio, Koch, Mordwilko, Tullgren, and Van
der Goot have been in constant use. The classification suggested by
Van der Goot ("Zur Systematik der Aphiden, " in Tijdschrift voor
Entomologie, vol. 56, p. 1913) has proved interesting, and although
the author has not felt at liberty to accept it in full, a translation of
his keys to the groups and genera has been included herewith (appen-
dix 1), which, it is hoped, will be of assistance in the making of
determinations.

CLASSIFICATION

The Aphididae belong to the order Homoptera, being closely
related to the Psyllidae, or jumping plant lice, the Aleyrodidae, or
white flies, and the Coccidae, or scale insects. The Aphididae, or
plant lice, are small, soft-bodied insects, ranging from less than one
to five or six millimeters in length. Typically there are four forms :
the apterous and the alate viviparous females, and the sexual forms,
the oviparous females and the males. There is considerable variation
from the above in different groups and species, as will be pointed
out under the discussions of the various species. The alate viviparous
females are the individuals most commonly taken by the collector and
the ones that usually show the best characters for determinations.
In the keys in this paper all characters refer to the alate viviparous
females (the alates) unless otherwise mentioned.

A SYNOPSIS OF THE APHIDIDAE

EXTERNAL ANATOMY

The body1 consists typically of three divisions, the head, thorax,
and abdomen. In the apterous forms the mesothorax and metathorax
are closely fused with the abdomen, while the prothorax and head
are distinct. In the alate forms the mesothorax and metathorax are
fused together and appear as a distinct division, the body appearing
to consist of four divisions, viz., the head, the prothorax, the meso-
thorax and metathorax, and the abdomen.

The head bears a pair of compound eyes, usually three ocelli, a
pair of three to six jointed antennae, and the beak. Of these, the
antennae show the best characters for determinations, not only of
species but of higher groups. They are either mounted on distinct
tubercles (Macrosiphini, certain Callipterini) or appear to arise from
the front of the head. They consist of from three to six segments,
the terminal one of which is usually provided with a projection or
spur. They are six-segmented in the Aphidinae (except Essigella and
Cerosipha), five- or six-segmented in the Pemphiginae (except in the
stem mothers of certain genera), and three-segmented in the Phyl-
loxerinae (except in Chermisina, in which the alate forms have five-
segmented and the sexual forms four-segmented antennae). The
spur of the terminal segment may be equal to or longer than the
segment (Aphidinae, in the Macrosiphini it attains its greatest length,
often being as much as ten times the length of the base) ; it may be
merely a short thumblike process (Pemphiginae, Lachnini, and cer-
tain Callipterini) ; or it may be apparently lacking (Phylloxerinae).
The two basal segments are always short, and quite regular in all
species. The remaining segments show the greatest diversity, par-
ticularly in number, size, and shape. Sensoria are always present on
some of the segments. There is one primary sensorium always present
at the distal end of the terminal segment, and when the antennae con-
sist of more than three segments, one also at the distal end of the
penultimate segment. These sensoria are fairly large and clear (some-
times furnished with a hairy fringe) and are more or less circular.
The accessory sensoria are a group of small indistinct sensoria, which
number from three to six, and which are located in close proximity

1 For a fuller discussion of the external characters consult the following
papers: Vickerey, E. A., A comparative study of the external anatomy of plant
lice, 12th Eept. Minnesota State Entomologist 1908 ; Sanborn, C. E., Kansas
Aphididae, Kansas Univ. Sci. Bull., vol. 3, 1904; Mordwilko, Alexander, Keys to
the groups and genera of the Aphididae, Ann. Mus. Zool. Imp. Acad. Sci. St.
Petersburg, vol. 13, pp. 362-364, 1908.

6 MISCELLANEOUS STUDIES

to the primary sensorium on the terminal segment. Secondary sen-
soria are usually present in the alate forms, but oftentimes absent
in the apterae of certain species. When present they are always on
the third segment, but in antennae consisting of five or six segments,
they may be present upon the fourth, fifth, and even sixth segments.
In the Pemphiginae they are arch-like or half rings, or form complete
rings about the segments. In the Aphidinae they are circular, oval,
or transversely linear, but are never rings or half rings. The shape
and number vary considerably, and are of specific importance. The
number may vary from as few as three or four (Myzocallis maureri
Swain), to as many as forty to fifty on the third segment, and many
also on the fourth and fifth (Myzus braggii Gillette). Unfortunately
these highly important characters were overlooked or not taken into
consideration by the earlier workers. The beak is four-jointed and
seems to arise from between the fore legs. It is always present (except
in the sexes of certain of the Phylloxerinae), but is seldom of specific
importance (except to distinguish Aphis bakeri Cowen from Aphis
senecio Swain, and in certain of the Lachini). It may be very short,
as in Aphis bakeri Cowen, where it reaches only slightly beyond the
first coxa, or it may be very long as in Stomaphis, where it is from
one and one-half to two times as long as the body. In leaf-feeding
species it is usually short, while in bark-feeding forms it is longer.
This is naturally necessary, for those that live on thick bark must have
a longer beak in order to reach through to the plant juices.

The thorax consists of three divisions, the last two of which are
usually more or less fused together, and considered as one ; the two
divisions being called, in this paper, the prothorax and the thorax.
On the lateral margins of the prothorax there is sometimes a pair of
small tubercles. These are not present in all species, however, and
they differ considerably in size in the various species. There are
three pairs of fairly long and slender legs (except in Phylloxerinae.
where the legs are greatly atrophied, approaching those of the
Coccinae in size). Typically the legs consist of four joints, the
coxa, the femora, the tibia, and the tarsus. In some genera the tarsi
may be atrophied (Atarsos, Mastapoda}. The comparative lengths
of the first and second segment of the tarsi are sometimes of generic
importance (Lachnini), and the comparative lengths of the hind tarsi
and the cornicles are oftentimes of specific importance (Aphis, Ptero-
comma). A small empodial hair is found between the claws in the
Aphidinae. In the Callipterina it is leaf-shaped or spatula-like. In

A SYNOPSIS OF THE APHIDIDAE 7

the Aphidina and Lachnina it is hair-like, usually being as long as the
claws (except in the Pterocommini, in which it is considerably
shorter than the claws). The wings are membraneous and hyaline
(except in certain Callipterini, Lachnini, and Macrosiphini), and are
held roof -like over the body when at rest (except Monellia, Phyllox-
erinae, Hormaphidina, in which they lie flat on the abdomen). The
veins of the fore wings are as follows: the costal and subcostal are
almost parallel with the anterior margin; the radial extends from
the posterior margin of the stigma to the outer margin of the wing,
being either curved or straight ; the discoidals, three in number, extend
from the subcostal to the posterior margin of the wing. The outer
or third discoidal (media, cubitus of some authors) may be simple
(Hormaphidina, Pemphigina), absent (Phylloxerinae), once-branched
(Schizoneurina), or twice-branched (Aphidinae, except Toxoptera).
On the anterior margin of the fore wing is a dusky spot located be-
tween the wing margin and the subcostal veins, and between the
distal ends of the costal and subcostal veins, known as the stigma or
Pterostigma. It is usually trapezoidal in shape, and does not extend
to the tip of the wing (except in Longistigma and Mindarus, in which
it reaches well beyond the tip of the wing). The hind wings have
one longitudinal and either one or two transverse veins.2 In the
Pemphiginae and Phylloxerinae dorsal wax glands are sometimes
present on the thorax, in which case their number, shape, and posi-
tion are of more or less specific importance.

The abdomen consists of nine more or less similar segments. The
coloration of the various segments, especially in species in which the
color is variegated, is sometimes of specific importance. In certain
species wax glands are present on the abdomen (Phylloxerinae, and
particularly the stem mothers of Pemphiginae) and may be of use
in making determinations. In the Aphidinae the presence or absence
and location of small lateral and dorsal tubercles are often important.
The anal segment consists of an anal plate and a cauda. The cauda
may not be separated from the abdomen (Pemphiginae, Lachnina), or
it may be short and conical (Aphidini), short and globular, being
constricted in the middle (Callipterina), or it may be long and
ensiform or sickle-shaped (Macrosiphini). The anal plate is usually
well rounded, being half-moon-shaped, or it may be emarginate or
bilobed (Callipterina). On the sixth (or fifth?) segment is a pair

2 For a full discussion of the venation see Patch, Edith M., Homologies of the
wing veins of the Aphididae, Psyllidae, Aleurodidae, and Coccidae, Ann. Entom.
Soe. Am., vol. 2, pp. 101-136, June, 1909.

8 MISCELLANEOUS STUDIES

of short tubular processes, the cornicles (honey tubes, nectaries of
some authors). These are quite valuable characters, both specific and
generic. In the Phylloxerinae and most of the Pemphiginae they are
lacking, but in the Aphidinae they are always present, and show a
great diversity of form. They may be merely pores (certain Callip-
terini, Cerosipha cupressi Swain, Lachnus taxifolia Swain), they may
be cylindrical, yet quite short (certain Callipterini, Chaitophorini) ;
they may be short and cylindrical or conical (Aphidini) ; they may
be truncate, cone-shape (Lachnini) ; they may be clavate and long
(certain Callipterina, Pterocommini, Macrosiphini) ; or they may be
long and cylindrical (particularly in Macrosiphum and Myzus).

BIOLOGY

Considerable variety is exhibited in the habits, life history, and
methods of reproduction, as well as in the structure and body form.
Reproduction is almost entirely parthenogenetic, although certain
species at certain times have a sexual reproduction. Fewer species
have sexual reproduction in California than in colder climates, due
to the fact that mild weather throughout the winter permits them to
live over, and hence the eggs are unnecessary. Many species produce
generation after generation parthenogenetically, and are most abun-
dant in the spring and early summer, but gradually disappear toward
midsummer, due partially to their predaceous and parasitic enemies,
and partially, undoubtedly, to the heat of the summer. Other species
regularly produce sexual forms in the fall, which lay eggs that hatch
the next spring. The forms hatching from the eggs are wingless
(except in Callipterini) and usually of a different form from the
later generations, and are known as the fundatfix or stem mother.
The fundatrix is always viviparous. Her progeny consists either of
all apterous or partly apterous and partly alate viviparous females
(fundatrigenia), which in turn produce other generations of funda-
trigeniae. The last asexual generation in the fall, which gives birth
to the sexual forms (sexuales), are known as sexupara, and are usually
alate. Oftentimes in the second or third and even fourth generation
there is a definite migration from one species of host plant to another,
where the aphids live over the summer (virgogenia), the sexupara
returning to the original species of host in the fall to give birth to
the sexuales, which lay their eggs there. Aphis malifoliue Fitch rep-
resents an example of this habit, the winter host being apple, the
summer plantain. Oftentimes the fall migrants (sexupara) of certain

A SYNOPSIS OF THE APHIDIDAE 9

species differ considerably in structure from the spring migrants
(fundatrigenia). This is particularly noticeable in the Pemphiginae.
Many species are confined throughout the season to one species of host,
others to one or two or a few species, while still others may live on
any of a number of hosts (Aphis senecio Swain, Rhopalosiphum per-
sicae ( Sulz. ) ) . All sustenance is derived from the plant juices of
the various hosts, but each species is usually confined more or less
definately to feeding on some certain part of the plant. Some live
entirely upon the leaves, some on the stems of the leaves and small
twigs, some on the trunks and larger branches, some on the roots,
some on the flower heads and racemes of the host, and still others
feed on almost any part of the plant. The greater number of species
are free living, but certain of the Aphidinae form pseudogalls (Aphis
pomi De Geer, Aphis malifoliae Fitch, Phyllaphis coweni (Cockerell) ),
while the Pemphiginae and Chermisina form true galls. Nearly all of
the Pemphigina spend at least part of the season on various species
of Populus, the Schizoneurina on TJlmus, while the Lachnini and
Chermisina are practically confined to the conifers. The Aphidinae
are found mostly on deciduous trees and herbaceous plants, although
some live on conifers (Myzaphis dbietinus (Walker), Nectarosiphon
morrisoni Swain).

ECONOMIC CONSIDERATIONS

From an economic standpoint most of the species are of no
importance, although there are many that are well known pests of
cultivated crops. For example the woolly apple aphis (Eriosoma
lanigera) is a world-wide pest of considerable importance to the apple.
The green and the rosy apple aphis (Aphis pomi, A. malifoliae) do a
large amount of injury in certain localities, and are extremely difficult
to control. The rose aphis (Macrosiphum rosae) is known the world
over, and although living unprotected and easily killed with any of
the common contact insecticides, it is recognized by everyone who has
grown roses in the dooryard as an extremely troublesome pest. The
walnut aphis (Chroniaphis juglandicola) , the cabbage aphis (Aphis
brassicae) , the green peach aphis or greenhouse aphis (Rhopalosiphum
persicae) are all well known pests. The common contact insecticides
are usually efficient for their control. Many species are kept well
in check by their predaceous and parasitic enemies, the ladybirds, the
syrphid flies, the lacewings, and the braconids. Of the ladybirds,
probably the most efficient in California are Coccinella californ-ica

10 MISCELLANEOUS STUDIES

Mann., Hippodamia convergens Guerin, and Scymnus nebulosus Le-
conte. Of the syrphid flies, those consuming the largest number of
aphids and the most abundant in the state3 are Catabomba pyrastrl
Osten-Sacken, Allograpta obliqua Say, Syrphus arcuatus Fallen, 8.
americanus Wied., 8. opinator Will., and Eupeodes volucris Osten-
Sacken. Chrysopa, calif omica Coq. and Sympherobius angustus Banks
are the most important aphid enemies among the lacewings. Among
the Braconidae there are two very common species in California,
Lysiphlebus testaceipes Cresson and Diaretus rapae Curtiss. Others
have been reared by the author and will be mentioned later. The
author wishes to thank Dr. L. 0. Howard and Mr. A. B. Gahan of
the Bureau of Entomology for their kindness in identifying the
various hymenopterous parasites of aphids sent to them.

SYNOPSIS
Family Aphididae Passerini

Passerini, Gli Afidi, 1860.

The family Aphididae Passerini is divided into three subfamilies
(following Alexander Mordwilko), which are: Aphidinae Buckton,
Pemphiginae Mordwilko, and Phylloxerinae Dreyfus. Van der Goot
considers but two subfamilies: Aphidinae v. d. G. and Chermisinae
v. d. G. His subfamily Aphidinae includes both the Aphidinae and
Pemphiginae of Mordwilko, while his Chermisinae is the same as
Mordwilko 's Phylloxerinae. Following is a translation of Van der
Goot 's descriptions of the two subfamilies :

Subfamily Aphidinae v. d. G. : Body very often without distinct groups of
glands for the secretion of wax. Antennae usually six- or seven-jointed [when
the terminal process of the sixth segment is longer than the segment he considers
it as the seventh segment]. Only in a few cases are the apterous forms with
three-segmented antennae. The primary sensoria usually have a distinct "haar-
kranz" [hairy fringe?]. Cornicles almost always and cauda often present. Fore
wings with four veins, the cubitus or media I very often divided: hind wings
usually with two cross-veins. Vivi-oviparous : the sexuales mostly of the usual
form.

Subfamily Chermisinae v. d. G. : Body almost always with distinct groups of
glands for the production of wax. Antennae three-segmented, often evidently five-
segmented. Sensoria always without "haarkranz. " Cornicles always absent.
Pore wings with three veins; hind wings with only one small vein. Always only
oviparous: sexuales dwarfish, with or without beak.

s Davidson, W. M., Syrphidae in California, Jour. Econ. Ent., vol. 9, pp. 454-
457, 1916.

A SYNOPSIS OF THE APHIDIDAE 11

The latter subfamily has been considered by the author as Phyllox-
erinae Dreyfus; the former as two subfamilies, Aphidinae Buckton
and Pemphiginae Mordwilko. Mordwilko gives the following char-
acters for these two subfamilies :

Subfamily Pemphiginae Mordw. : Antennae of the alate forms five- or six-
segmented, the third bearing a specifically definite number of transverse or arch-
like sensoria; short, usually not longer than the head and thorax. The apterous
parthenogenetic females have four- to six-segmented antennae, but these are
sometimes reduced to three or even to two segments. The fore wings of the
alate forms have four transverse veins, of which the third or cubital vein [third
discoidal] is either simple or once-branched. The hind wings have one or two
transverse veins. The cornicles are either entirely absent or very slightly devel-
oped, and in the latter case may not be present in all the forms of one species.

Subfamily Aphidinae Buckton: Antennae always six-segmented, except in the
stem mother of some species, and in the genus Sipha Passerini. [This genus is
not represented in California. In Essigella Del Guercio, Cerosipha Del Guercio,
and Trifidaphis Del Guercio, three Californian genera described since the publi-
cation of Mordwilko 's paper, the antennae are but five-segmented.] The last
antennal segment often ends in a long thread-like filament which may be longer
than the segment. Antennae with a long filament are mostly from half the
length of the body to longer than the body. The antennal filament is character-
istic only for this subfamily; some genera of the groups Lachnina and Callipterina
have a very short filament, and the antennae are not longer than the head and
thorax. The sensoria are small and are shaped like dots, circles, or transverse
holes, but never archlike or half-rings. Segment 3 bears the largest number,
especially in the alate forms. The cubitus [third discoidal] of the fore wings is
usually twice-branched although there are some exceptions, as Toxoptera Koch.
Most species have long cylindrical cornicles which are often clavate in the middle.
Sometimes they may be greatly reduced or poorly developed, and, as in Lachnina
and Callipterina, they may be replaced by cupola-shaped elevations. A cauda
is usually present, being conicle, ensiform, or globular, although in Lachnina it is
not evident. The sexual forms have beaks, and become quite large.

Subfamily Aphidinae Buckton

Buckton, Mono, British Aphides, 1883.

This subfamily is divided into three groups, following Carl Borner
(Sorauer, Paul, Handbuch der Pflanzenkrankheiten, vol. 3, p. 664,
1913). Borner considers the family Aphididae as a superfamily, and
divides it into four families ; so this subfamily Aphidinae he considers
a family, and the various groups as subfamilies. Below is a trans-
lation of his key :

1. Claws with spatula-like or leaf -shaped empodial hairs (fig. 1). Cornicles vari-
ously formed, bare. Pubescence of larvae as in Aphidina. The majority
of the species live free and monophagous on trees, only seldom on herbaceous

plants, and never migrate collectively Group Callipterina

— Claws with simple empodial hairs (fig. 2), often hard to see 2

12 MISCELLANEOUS STUDIES

2. Antennae with short terminal joint (fig. 3), (except in Pterocommini, but then
the cauda is not tail-like). Body ridges with more than six longitudinal
rows of hairs. Hairy covering mostly thick. Cauda not lengthened tail-like,
anal plate widely rounded (fig. 5). Wax glands either present or lacking.
Mostly strongly monophagous forms, at times of remarkable size. Found

mostly on tree growths and without change of hosts Group Lachnina

— Terminal joint of antennae always with a long, slender filamentous projection
(fig. 4). Body ridges of young larvae at most with only six longitudinal
rows of hairs, which may be increased after the first molt. Cauda either
short or lengthened tail-like, anal plate widely rounded (fig. 6). Species
monophagous or polyphagous, many with a change of host plants. On
tree or herbaceous growths Group Aphldina

Group Callipterina Mord
(Subfamily Callipterinae Borner)

Mordwilko, Ann. Imperial Acad. Sci., St. Petersb., 1908.

Borner, in Sorauer, Handbuch der Pflanzenkrankheiten, vol. 3, p. 664, 1913.

According to Borner this group consists of two tribes, -the Phyl-
laphidini and the Callipterini. He divides the Callipterini into two
groups, the Callipterini and the Chaitophori. The author has followed
him to a certain extent, but has given each of the last two groups equal
rank with the Phyllaphidini, and thus considers this group, Callip-
terina, as consisting of three tribes. Below is a key to the same :

1. Wax glands with faceted pore fields present. Antennae as in Lachnina

(fig. 13). Pubescence delicate Tribe Phyllaphidini

— Wax glands lacking or without faceted pore fields. Pubescence often very
remarkable. Terminal joint of the antennae often lengthened into a bristle
(fig. 30) 2

2. Anal plate more or less emarginate or bilobed (fig. 7), except in Euceraphis

Koch Tribe Callipterini

- Anal plate widely, truncate or rounded (fig. 8) Tribe Chaitophorini

Tribe Phillaphidini Borner

Borner, in Sorauer, Handbuch der Pflanzenkrankheiten, vol. 3, p. 664, 1913.

This tribe Phyllaphidini consists of but one genus, Phyllaphis
Koch, which is represented in California by three species.

L Genus Phyllaphis Koch
Koch, Die Pflanzenlause, p. 248, 1857. Type Aphis fagl Linn.

KEY TO CALIFORNIA SPECIES

1. Alate viviparous females unknown. Wing venation of alate males similar to
that of Eriosoma spp. (fig. 17). Forming pseudogalls on edges of leaves
or living free in masses of white flocculence on leaves of Quercus spp.

quercicola Baker

A SYNOPSIS OF THE APHIDIDAE 13

— Alate viviparous females common. Venation normal, the third discoidal being

twice-branched. Not on Quercus spp 2

2. Antennae short, stout, with oval transverse sensoria (fig. 13). Forming galls
on Arctostaphylos spp. (and Arbutus spp.) _ coweni (Ckll.)

— Antennae longer and narrower with circular sensoria (figs. 9, 14-17). Living

under thick masses of white flocculence on Fagus spp fagl (Linn.)

1. Phyllaphis coweni (Ckll.)

Figure 13

Cockerell, Can. Ent., vol. 37, pp. 391-392. 1905. Pemphigus (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 4, pp. 559, 1911. Cryptosiphum tahoense

n.sp. (desc.).

Davidson, Jour. Econ. Ent., vol. 5, p. 404, 1912 (list).
Essig, Pom. Jour. Ent. Zool., vol. 7, pp. 187-195, 1915 (desc.).
Records. — Arctostaphylos manzanita; Oakville, Napa County, February, 1913
(E. L. Brannigan) ; Mount Diablo, Contra Costa County (Davidson) ; Jasper
Eidge, Santa Clara County, October, 1914 (E. A. Cornwell) ; Pine Hills, San Diego
County, June, 1916.* A. pumella, A. tomentosa, Lake Tahoe, August, 1911
(Davidson) : A. glauca, Alpine, San Diego County, June, 1916.

This species is found more or less abundantly throughout the state
wherever its host plants occur. Essig (1915) states it is found
throughout the Rocky, Sierra Nevada, and Coast Range mountains,
being more abundant in the central and northern parts of the state.
The author has found it to be extremely abundant in the Cuyamaca
and Laguna mountains in the extreme southern part of the state.
The insects can be found at any time of the year in the galls on
manzanita although most abundantly in the early fall. Collections
by the author in June showed that the stem mothers and young vir-
gogeniae only were present. A few weeks later the alate females were
abundant, while in August the sexual es begin to appear. However,
the alate viviparous females have been found in October and in
February. This species forms galls on the leaves, and flower and
fruit stalks of its host. Usually there is but one gall to a leaf,
although sometimes four or five may be found. When first formed
these galls are concolorous with the leaves; but as they become older
they turn more and more reddish in color, until when mature they are
a very bright red.

2. Phyllaphis fagi (Linn.)

Figures 9 to 12

Linnaeus, Syst. Nat., vol. 2, p. 735, 1735. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).

Eecords. — Fagus sp., Palo Alto, 1910 (Davidson) ; Fagus sylvatica, Stanford
University, April to May, 1915.

* Records in which no collector 's name is mentioned refer to collections made
by the author.

14 MISCELLANEOUS STUDIES

This species has been taken only in the vicinity of Stanford Uni-
versity, where it infests copper beach (Fagus sylvatica). It may be
easily recognized by the masses of whitish flocculence on the under
side of the leaves. Each mass contains one individual, which is
entirely hidden by it. In looking up the literature of this species
the author found that there has been no description of it published in
America, so below is included a brief description of specimens taken
near Stanford University on April 28 and May 29, 1915.

Alate viviparous female. — Prevailing color dark green, covered
.with a whitish flocculence. This flocculence consists of wax threads
as much as 3 mm. long. Head dusky, with frontal margin black.
Eyes red. Antennae dusky, except II and basal one-third of III,
which are pale. Beak pale with apex and joints dusky. Thorax
dusky green with lobes black. Abdomen dark green with a row of
black spots on each margin and about seven black transverse dorsal
bands. Cornicles black. Cauda and anal plate concolorous with
abdomen with distal margins slightly darker. First and second femora
pale with apices only dusky; third femora dusky throughout. First
tibiae pale with apex dusky; second and third dusky throughout.
Tarsi black. Wings hyaline, stigma gray.

Head twice as wide as long, furnished with many small wax
glands. Antennae reaching to the cornicles or to the base of the
cauda, set on small tubercles (fig. 12). Ill is the longest segment,
followed by IV, V, and VI. VI spur is merely a thumb-like projec-
tion (fig. 16). The usual primary and accessory sensoria are present
on V and VI. Secondary sensoria are found only on III (fig. 9).
These are fairly large, almost circular, and placed in a single row
along the segment. They number from four to seven, five being the
average. The beak is short, reaching but slightly beyond the first
coxae. The wings are normal, with a twice-branched third discoidal.
The cornicles are merely small pores. The cauda is short and knobbed,
the anal plate emarginate or bilobed (fig. 11).

Measurements : Body length 2.0 to 2.4 mm., width 0.8 to 1.04 mm.,
antennae total 1.55 to 2.06 mm., Ill 0.591 to 0.77 mm., IV 0.34 to 0.47
mm., V 0.27 to 0.39 mm., VI 0.19 to 0.25 mm., cornicles (diameter)
0.05 mm.

Apterous viviparous female. — Prevailing color under flocculence
pale yellowish green. Light brown markings as follows: two rows
of four spots each across the prothorax, one large spot on each margin
and one on the dorsum of the thorax, four spots on each abdominal

A SYNOPSIS OF THE APHIDIDAE 15

segment, two dorsal and two marginal. Antennae pale except VI,
apical two-thirds of V, and apical one-third of IV. Legs pale with
light brown spots at joints; tarsi black. Cauda small and conicle,
cornicles not evident.

Measurements : Body length 2.9 to 3.0 mm., width 0.96 to 1.2 mm.,
antennae total 1.26 mm., Ill 0.36 mm., IV 0.32 mm., V 0.204 mm.,
VI 0.205 mm.

3. Phyllaphis quercicola Baker

Figures 14 to 20

Clarke, Can. Ent., vol. 35, p. 248, 1903. Schizoneura querci (Fitch) (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. S. querci (Fitch) (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. S. querci (Fitch) (list).
Davis, Ent. News, vol. 22, p. 241, 1911. Phyllaphis querci (Fitch) (biblig.)
Davidson, Jour. Econ. Ent., vol. 7, p. 127, 1914. P. querci (Fitch) (note).
Gillette, Ent. News, vol. 25, p. 274, 1914. Phyllaphis sp. (list).
Baker, Ent. News, vol. 27, p. 362, 1916. P. quercicola n.n. for P. querci
(Fitch) of Davis.

Eecords. — Quercus agrifolia; Placer, Contra Costa, Santa Clara counties
(Davidson, Clarke); Stanford University, April, 1915; Berkeley, September, 1915;
Wynola, San Diego County, June, 1916; Charter Oak, Los Angeles County, Novem-
ber, 1916. Quercus lobata, Stanford University (Davidson) ; Q. wislizenii, Placer
County (Davidson); Q. dumosa, San Diego, August, 1915; Quercus sp., Spreckels,
Monterey County, 1913 (Gillette).

This is a very common species of woolly aphis on the oaks, par-
ticularly the live oak, through southern and central California.
According to Davidson (1914) the stem mothers occur in pseudogalls
on the edges of the leaves. The second generation lice, when mature,
leave these galls to live on the upper and lower surfaces of the leaves,
unprotected except for their -woolly covering. The sexes, apterous
oviparous females and alate males, occur late in the fall. The vivi-
parous generations are all apterous. The writer has observed the
stem mothers as late as August in San Diego County, while he has
found the viviparous females on the under side of the leaves as early
as mid-June in Berkeley.

The identity of this species has never been definitely established.
It was thought to be the species described by Fitch (Kept. Ins. N. Y.,
vol. 5, p. 804, 1859) as Eriosoma querci, but in 1916 Baker pointed
out the identity of Eriosama querci Fitch, proving it to be identical
with a species of Anoecia found on Cornus and formerly considered
to be A. corni Fab. Baker's decision is that the Quercus-Cornus
species of the eastern United States is Anoecia querci (Fitch) and

16 MISCELLANEOUS STUDIES

is distinct from our western one. In 1911 Davis described a species
of woolly aphis from oak under the name of Phyllaphis querci (Fitch)
stating that it is the same one as listed by Davidson. Baker proposes
the name Phyllaphis quercicola for this species described by Davis.
Consequently it is so listed in this paper. This species is not a typical
Phyllaphis, but it fits that genus better than any other so is placed
there provisionally. The figures (14—20) are from a specimen of alate
male in the Davidson collection in Stanford University.

Tribe Callipterini Wilson

Wilson, Can. Ent., vol. 42, p. 253, 1910.

The genera included in this tribe differ somewhat as considered by
various entomologists. Since Wilson has worked out the synonomy
of the various genera very well he is followed in preference to some
of the European authors, although there are some points in which
he is mistaken. For instance, he places Pterocalli-s Passerini, Callip-
teroides Mordwilko, Tuberculatus Mordwilko, Subcallipterus Mord-
wilko, and Therioaphis Walker as synonyms of Myzooallis Passerini.
In regard to this, he states, "In 1894 Mordwilko used A. coryli Goetze
as the type of his genus Callipteroides, but as this species ..." He
is mistaken in this, for in the paper referred to, Mordwilko used
A. coryli Goetze as the type of the genus Myzocallis Passerini, and in
1908 he gave as the type of Callipteroides, Callipterus nigritarsus
Heyden (betulae Koch). If nigritarsus Hey den is a synonym of
betulae Koch, as Mordwilko indicates, then Callipteroides is a synonym
of Enceraphis Walker, for C. betulae Koch certainly falls into this
genus, as described by Wilson himself. The key to the California
genera below is adapted from Wilson's key (Can. Ent., vol. 42, pp.
253-254, 1910).

KEY TO CALIFORNIA GENERA OF CALLIPTERINI

1. Antennal tubercles prominent (fig. 21); antennae always exceedingly long.... 2

— Antennal tubercles wanting or very small (fig. 22) ; antennae variable, some-

times shorter than the body 3

2. Cornicles very long and large (figs. 23-24) 4

— Cornicles very short and more or less constricted in the middle 5

— Cornicles little more than pores (fig. 25). Wings held horizontal at rest.

Monellia Oestlund

3. Cornicles distinct, usually being longer than broad in the middle (fig. 26) 6

— Cornicles little more than pores, and broader than long (fig. 25). Wings

held horizontal at rest ... ....Monellia Oestlund

A SYNOPSIS OF THE APHIDIDAE 17

4. Cornicles one-fourth the length of the body or more, swollen in the middle

(fig. 24) Drepanosiphum Koch

— Cornicles large and nearly one-fourth the length of the body, swollen at the

base and tapering toward the middle (fig. 23) ....Drepanaphis Del Guercio

5. Inner side of antennal tubercles about one-half the length of the inner side

of the first antennal joint (fig. 29) Euceraphis Walker

— Inner side of antennal tubercles more than one-half the length of the inner

side of the first antennal segment (figs. 27-28) Calaphis Walsh

6. Antennae longer than body, except in Callipterinella, with VI spur not much

shorter than VI base (fig. 31) 7

— Antennae shorter than the body, with VI spur very short, often being little

more than a nail-like process (fig. 34) 9

7. VI spur considerably longer than VI base, being one and one-half to two

times as long. Anal plate emarginate but not deeply bilobed.

Callipterinella Van der Goot

— VI spur about equal to or shorter than VI base. Anal plate deeply bilobed 8

8. VI spur and VI base subequal (fig. 31). Cornicles twice as long as broad

in the middle and constricted in the middle (figs. 26, 32).

Myzocallis Passerini

— VI spur shorter than VI base (fig. 30). Cornicles much broadened at base

(fig. 33) Eucallipterus Schouteden

9. VI spur less than one-half the length of VI base (fig. 34). Cornicles not

longer than broad at the base, and constricted in the middle (fig. 35).

Chromaphis Walker

— VI spur at least one-half as long at VI base (figs. 63, 66). Cornicles short,

about as long as broad and placed on a broad base Callipterus Koch

2. Genus Drepanosiphum Koch

Koch, Die Pflanzenlause, p. 201, 1855. Type Aphis palantanoides Schrank.

4. Drepanosiphum platanoides (Schrank)

Figures 21, 24, 36

Schrank, Fauna Boic., vol. 2, p. 1206, 1801. Aphis (orig. desc.).
Wilson, Jour. Econ. Ent., vol. 2, p. 349, 1909 (desc. ala. vivi., ala. ovi.

females).

Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 759, 1912 (list).

Records. — Acer macrophyllum, A. negundo; Berkeley, 1915 (Essig) ; Stan-
ford University, October, 1914, April, 1915; A. pseudoplatanus, Stanford Univer-
sity, November, 1914 (Morrison); A. saccharum, Berkeley, June, 1915; Platamis
racemosus, Stanford University (Davidson) ; Acer sp., San Lorenzo, 1908 (Wil-
son).

This is a very common species in the San Francisco Bay region
on various species of maples, and on box elder and western sycamore.
In April the alate and apterous viviparous females are abundant,
remaining so throughout the summer and early fall. In the later
fall (October and November) the sexes appear. Just where the eggs

18 MISCELLANEOUS STUDIES

are laid the author is unable to say. A curious fact is that thp
oviparous females are alate as well as apterous. The author has never
seen the alate forms, but Wilson (1908) describes them.

3. Genus Drepanaphis Del Guercio

Del Guercio, Eivista di patologia vegetable, vol. 4, pp. 49-53, 1909. Type
Siphonophora acerifolii Thomas.

5. Drepanaphis acerifolii (Thomas)

Figures 23, 37

Thomas, Illinois Lab. Nat. Hist., Bull. 2, p. 4, 1878. Siphonophora (orig.

desc. ) .

Clarke, Can. Ent., vol. 35, p. 249, 1903. Drepanosiphum (list).
Sanborn, Kan. Univ. Sci., Bull. 3, p. 45, 1904. Drepanosiphum (desc. ala.).
Davidson, Jour. Econ. Ent., vol. 2, p. 303, 1909. Drepanosiphum (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910. Macrosiphum (list).
Essig, Pom. Jour. Ent., vol. 4, p. 760, 1912 (list).
Essig, Mon. Bull., Cal. Comm. Hort., vol. 3, p. 85, 1914 (list).
Essig, Mon. Bull., Cal. Comm. Hort., vol. 3, p. 445, 1914 (list).

Records. — Acer sp. : Stanford University (Davidson); Sacramento (Essig);
Hanford, Fresno County (B. V. Sharp) ; A. macrophyllum, A. saccharinum, Berke-
ley, July to October, 1915; Riverside, October, 1916; A. dasycarpum, A. plat-
anoides, Berkeley, 1915 (Essig); Quercus sp. (live oak), Berkeley (Clarke) (f).

This is as common a species on maple in the San Francisco Bay
region as the preceding one. It has also been taken in the Sacramento
and the San Joaquin valleys, and in southern California. It is a
species easily recognized by its dark markings and the dorsal tubercles
on the first and second abdominal segments.

4. Genus Calaphis Walsh

Figure 28
Walsh, Proc. Ent. Soc. Phila., vol. 1, p. 301, 1863. Type C. betulella n.sp.

6. Calaphis betulaecolens (Fitch)

Figures 27-38

Fitch, Cat. Homop. N. Y., p. 66, 1851. Aphis (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 249, 1903. Callipterus (list).

Davidson, Jour. Econ. Ent., vol. 2, p. 301, 1909. Callipterus (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus (list).

Essig, Pom. Jour. Ent., vol. 3, p. 556, 1911 (syn.).

Davidson, Jour. Econ. Ent., vol. 5, p. 404, 1912 (desc. sexes).

Essig, Pom. Jour. Ent., vol. 4, p. 760, 1912 (list).

Essig, Mon. Bull., Cal. Comm. Hort., vol. 3, p. 445, 1914 (list).

Baker, Proc. Ent. Soc., Washington, vol. 18, p. 186, 1916 (desc.).

Records. — Betula, sp., Alameda, Contra Costa, Santa Clara counties (Clarke,
Davidson, Essig, Morrison, and the author).

A SYNOPSIS OF THE APHIDIDAE 19

This is a common species of aphid on birch (Betula spp.) in the
San Francisco Bay region. In the early part of March the eggs begin
to hatch. In 1915 at Stanford University eggs began to hatch on
March 8, the process continuing for several days. A month later
both alate and apterous females were quite abundant; the alate
females being undoubtedly the stem mothers, the apterae belonging
to the second generation. Viviparous generations appeared through-
out the summer. During August the sexes, alate males and apterous
oviparous females, occurred. In 1914 the sexes and sexupara were
noticed on August 28. Egg laying occurred shortly afterward, the
eggs being laid in the crotches of the twigs and under the curled
edges of the bark. Birch is the only recorded host plant.

5. Genus Euceraphis Walker

Walker, The Zoologist, p. 2001, 1870. Type Aphis letulae Koch.
KEY TO CALIFORNIA SPECIES

1. Body light green; third joint of antennae with about 13-18 sensoria on basal
one-half (fig. 39) giUettei Dvdn.

— Body yellow with dark markings on head and thorax, and often with as
many as eight black transverse stripes on the abdomen (the number varies
between none and eight) ; third antennal segment with 19-25 sensoria
(fig. 40) betulae (Koch)

7. Euceraphis betulae (Koch)

Figures 29, 40

Koch, Die Pflanzenlause, p. 217, 1855. Callipterus (orig. dese.).
Davidson, Jour. Econ. Ent., vol. 5, p. 405, 1913 (desc. ovi. female).
Davidson, Jour. Econ. Ent., vol. 7, p. 129, 1914 (desc. stem mother).

Records. — Betula sp. : Oakland (Davidson); Palo Alto, March to April, 1915.

Davidson lists this species, describing the stem mother and
oviparous female from the San Francisco Bay region. The author
found it in Palo Alto during March and April, 1915, on Betula alba.
According to Davidson the stem mothers hatch from the eggs about
the middle of February, feeding on the stems until the leaves open in
March. The viviparous generations occur during the summer. He
took the oviparous females in November. His description of the stem
mother gives three dusky transverse bands on the abdomen. The
author has found this to be variable, the number ranging from none
to eight.

20 MISCELLANEOUS STUDIES

8. Euceraphis gillettei Davidson

Figure 39

Clarke, Can. Ent., vol. 35, p. 248, 1903. Lachnus alnifoliae Fitch (list).
Davidson, Jour. Econ. Ent, vol. 2, p. 300, 1909. L. alnifoliae Fitch (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 375, 1910. L. alnifoliae Fitch (list).
Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912. Lachnus alnifoliae Fitch

(list).

Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912. Lahnus alnifoliae Fitch (note).
Davidson, Jour. Econ. Ent., vol. 8, p. 421, 1915 (orig. dese.).

Records. — Alnus rhombifolia; Berkeley (Clarke), Stanford University, San
Jose, Walnut Creek (Davidson), Stanford University, March, 1915.

This species was reported from alder by Clarke and Davidson as
Lachnus alnifoliae Fitch. Essig, in his Host plant list of California
Aphididae, lists Callipterus alnifoliae (Fitch) on Alnus rhombifolia,
but later states that this citation should be Lachnus alnifoliae Fitch.
Therefore he referred to this new species of Davidson. The author
took both apterous and alate viviparous females of this species on
Alnus rhombifolia, along the banks of the San Francisquito Creek,
near Stanford University, on March 19, 1915. During the spring it
was quite common there.

6. Genus Eucallipterus Schouteden

Schouteden, Mem. Soc. Ent. Belg., vol. 12, 1906. Type Aphis tiliae Linn.
KEY TO CALIFORNIA SPECIES

1. Wings hyaline; III pale except at the apex, with 5-7 sensoria on the basal

one-fifth (fig. 41) flava (Dvdn.)

2. Wings with veins clouded; III with apical one-fifth and basal one-half dusky,

and with about 13-15 sensoria on the basal one-half (fig. 42).

tiliae (Linn.)

9. Eucallipterus flava (Davidson)

Figure 41

Davidson, Jour. Econ. Ent., vol. 5, p. 406, 1912. Euceraphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 8, p. 423, 1915 (desc. sexes).

Records. — Alnus rhombifolia; San Jose, Walnut Creek (Davidson), Stanford
University (Morrison).

This is an uncommon species in the San Francisco Bay region on
Alnus rhombifoli-a, occurring on the under side of the leaves. The
author has never collected it, but has specimens from Davidson, taken

A SYNOPSIS OF THE APHIDIDAE 21

in April, 1913, near Walnut Creek, Contra Costa County. According
to Davidson the sexes appear in October, egg laying occurring during
the first part of November. The eggs are laid at the axils of the new
buds and on the twigs or canes. These hatch the following spring, the
stem mothers being found in the early part of April.

10. Eucallipterus tiliae (Linn.)

Figures 7, 30, 33, 42, 50

Linnaeus, Syst. Nat., vol. 2, p. 734, 1735. Aphis (orig. dese.).
Davis, Ann. Ent. Soc. Amer., vol. 2, p. 33, 1909. Callipterus (desc., biblio.).
Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909. Callipterus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910. Callipterus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 763, 1912 (list).

Records. — Tilia americana, Tilia curopea; Stanford University (Davidson),
Berkeley, August, 1914 (Essig) ; Stanford University, April to May, 1915; Berke-
ley, June, 1915.

In the Sari Francisco Bay region this very pretty aphid is quite
common on basswood or linden. The author has taken it throughout
April, May, and June. Essig found it abundantly in August. It is
very easily recognized when found at rest on the under side of the
leaves of its host by the two black lines extending from the front o£
the head along the margins of the thorax and joining with the costal
margins of the wings. It so appears that these lines are continuous
from the front to the tip of the wings.

7. Genus Myzocallis Passerini

Passerini, Gli Afidi, p. 28, 1860. Type Aphis coryli Goetze.
KEY TO CALIFORNIA SPECIES

1. Wings hyaline 6

— Wings not hyaline, with portions shaded (fig. 262) .. 2

2. Costal cell of wings hyaline (figs. 266, 267) 3

— Costal cell of forewings dusky or shaded (figs. 263, 264) 5

3. First discoidal vein dusky, otherwise the wing is hyaline. VI with spur shorter

than base. Apical one-half of III dusky (fig. 47). Cornicles pale. Found
on Alnus spp alnifoliae (Fitch)

— Wings not as above (figs. 266, 267). VI with spur either equal to or longer

than base. Ill with less than apical one-half dusky 4

4. Cornicles pale. Abdomen without dusky dorsal markings. On Quercus spp.

maurerl Swain

— Cornicles dusky (fig. 62). Abdomen with dusky dorsal markings. On Casta-

nea spp. and Quercus spp davidsoni Swain

22 MISCELLANEOUS STUDIES

5. Cornicles pale. Wings with greater portion cloudy (fig. 262). Antennae with

only tips of III to VI dusky. On Quercus spp discolor (Monell)

— Cornicles pale with apex dusky. Wings with dusky band along costal margin

(fig. 263). Antennae with tips of III and IV, apical one-half of V, and
all of VI and spur dusky. On Quercus spp bellus (Walsh)

6. Abdomen with four spine-like tubercles on the dorsum of the first segment.

VI with base and spur subequal, III being considerably longer than both.
Cornicles pale, small, and inconspicuous. On Ulmus spp.

ulmifolii (Monell)
-•— Abdomen without tubercles as above 7

7. Ill shorter than VI (base and spur). On Quercus spp punctatus (Monell)

— Ill not shorter than VI (base and spur) 8

8. VI with spur about twice as long as base (fig. 44). Cornicles pale. On

Corylus spp coryli (Goetze)

— VI with spur at most only slightly longer than base 9

9. Ill with apex only dusky (figs. 57, 58). Cauda pale 10

- Ill dusky throughout (fig. 268) or with apex and a band near the base dusky

(fig. 48). Cauda dusky 11

10. Cornicles pale. Antennae longer than body. Sensoria on III (two or three

in number) small and located close to the base of the segment (fig. 57), On
Pasania spp pasaniae Dvdn.

— Cornicles dusky, at least apical one-half. Antennae not longer than the body.

Sensoria on III (five or more in number) fairly large and on basal two-
thirds of segment (fig. 58). On Quercus spp quercus (Kalt.)

11. Abdomen with dusky dorsal markings. Ill dusky throughout (fig. 268). On

Arundo spp arundinariae Essig

— Abdomen without dusky dorsal markings. Ill with apex and band near base

dusky (fig. 48). On Arundo spp arundicolens (Clarke)

11. Myzocallis alnjfoliae (Fitch)

Figure 47

Fitch, Cat. Homop. N. Y., p. 67, 1851. Lachnus (orig. desc.).
Essig, Pom. Jour. Ent., vol. 4, p. 764 (762), 1912. M. alni (Fabr.) (desc.

viviparae).

Baker, Jour. Econ. Ent., vol. 10, p. 421, 1917 (note).
Records. — Alnus rhombifolia; Santa Paula (Essig).

Only once has this species been taken in California, by Essig in
August, 1911, near Santa Paula, Ventura County. At that time it
was very abundant on the under side of the leaves, causing a large
amount of sooty mold.

32. Myzocallis arundicolens (Clarke)

Figures 22, 48, 51, 52

Clarke, Can. Ent., vol. 35, p. 249, 1903. Callipterus (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 301, 1909. Callipterus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 762, 1912 (list, in part).
Essig, Univ. Calif. Publ. Entom., vol. 1, p. 305, 1917 (desc.).

H

A SYNOPSIS OF THE APHIDIDAE 23

Records. — Bamboo, Berkeley (Clarke, Essig) ; Arundinaria japonica, Berkeley,
June, 1915.

In the San Francisco Bay region and in the Sacramento Valley
this species is often found infesting the upper and lower surfaces of
the leaves of various bamboos, particularly species of Arundinaria,
Bambusa, and Phyllostachys, and the giant reed (Arundo donax).
Reports list it from Alameda, Sacramento, San Francisco, and Santa
Clara counties. The species described by Davidson (1914) as Eucal-
lipterus arundicolens (Clarke) and reported from southern California
by Essig (1912) proves to be distinct, and was described by Essig
(1917) as M. arundinariae. The following brief description is from
a collection made by the author on June 9, 1915, from Arundinarm
japonica on the campus of the University of California in Berkeley.

Alate viviparous female. — (Second generation?) Prevailing
color, pale yellow. Head twice as wide as long, pale yellow, with
prominent red eyes. Antennal tubercles absent. Antennae longer
than body ; formula III, IV, V, VI spur, VI base, I, II. Segments all
pale except the margins of I and II, the apices of III, IV, V, and a
band about one-sixth the length of III a short distance from the base
of III (fig. 48), which are black, and VI which is slightly dusky.
There are five or six transverse secondary sensoria on III, located in
the dark band. The usual primary sensoria are present on V and VI,
and the usual accessor sensoria on VI. Beak pale and short, reaching
only to the middle of the first coxae. Thorax and abdomen normal,
pale yellow, without tubercles or dusky markings. Cornicles (fig. 51)
pale, short, broader at base than at apex. Cauda short, constricted
in the middle, with distal end black. Anal plate (fig. 52) pale, deeply
bilobed. Wings normal, hyaline, with the first and second discoidal
veins and the base of the stigmal vein darker than the others. There
is a perceptible shading at the tip of each vein.

Measurements: Body length 1.326 to 2.023 mm. (av. 1.644 mm.),
width of thorax 0.51 to 0.68 mm. (av. 0.612 mm.), antennae total
2.839 to 3.077 mm. (av. 2.9299 mm.), Ill 0.8925 to 0.986 mm. (av.
0.9324 mm.), IV 0.578 to 0.663 mm. (av. 0.6423 mm.), V 0.527 to
0.561 mm. (av. 0.5403 mm.), VI base 0.306 to 0.323 mm. (av. 0.3103
mm.), VI spur 0.34 to 0.425 mm. (av. 0.3691 mm.), cornicle 0.595 to
0.765 mm. (av. 0.7002 mm.), cauda 0.153 mm., wing length 2.25 to
3.96 mm. (av. 2.9097 mm.), width 1.02 to 1.122 mm. (av. 1.071 mm.),
expansion 6.341 to 6.97 mm. (av. 6.6555 mm.).

24 MISCELLANEOUS STUDIES

13. Myzocallis arundinariae Essig

Figure 268

Essig, Pom. Jour. Ent., vol. 4, p. 762, 1912. M. arundicolens (Clarke) (in
part).

Davidson, Jour. Econ. Ent., vol. 7, p. 129, 1914. Eucallipterus arundi-
colens (Clarke) (desc. viviparae).

Essig, Univ. Calif. Publ. Entom., vol. 1, pp. 302-305, 1917 (orig. desc.).

Becords. — Arundo sp., San Francisco Bay region (Davidson) ; Arundinaria
japonica, Santa Barbara (Essig); Riverside, January to May, 1917; Arundo
donax, San Diego, April to June, 1916.

This is the commonest bamboo-infesting species in southern Cali-
fornia and parts of central California. For some time it was con-
sidered as M. arundicolens (Clarke) but this past year Essig pointed
out the differences, describing it as a new species.

14. Myzocallis bellus (Walsh)

Figures 45, 46

Walsh, Proc. Ent. Soc. Phila., vol. 1, p. 299, 1862. Aphis (orig. desc.).
Essig, Pom. Jour. Ent. Zool., vol. 7, pp. 195-200, 1915. Callipterus (desc.).

Eecords. — Quercus agrifolia, Alhambra, Los Angeles County (Essig) ; Ventura
(Essig).

Two collections have been made of this species in California, both
in southern California, in January, 1912, in Alhambra, and in May,
1913, in Ventura. Both of these consisted only of the alate females
(stem mothers), and were described by Essig.

15. Myzocallis davidsoni Swain

Figures 60, 61, 62, 267

Clarke, Can. Ent., vol. 35, p. 249, 1903. Callipterus castaneae Fitch (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus castaneae
(Buckton) (list).

Davidson, Jour. Econ. Ent., vol. 5, p. 405, 1912. Calapliis castaneae (Buck-
ton) (desc. sexuales).

Essig, Pom. Jour. Ent., vol. 4, p. 760, 1912. Calaphis castaneae (Fitch)
(list).

Swain, Trans. Am. Ent. Soc., vol. 44, p. 1, 1918 (orig. desc.).

Eecords. — Castanea sp., Berkeley (Clarke, Essig, Swain), Stanford University
(Davidson, Swain), San Jose (Davidson); Quercus pedunculata, Berkeley (Swain,
Essig).

This species was first reported in California by Clarke as Callip-
terus castaneae Fitch and later by Davidson as Cattipterus castaneae

A SYNOPSIS OF THE APHIDIDAE 25

Buckton. Recently the author described the species from specimens
taken in Berkeley on chestnut and oak. It cannot be the Callipterus
castaneae of Fitch, because the latter is really a Calaphis. It may be
the same species that Buckton had when describing his Callipterus
castaneae, in which case his name would be dropped as Fitch's species
has priority, and is replaced by the author's name, M. davidsoni. It
is more or less common throughout the San Francisco Bay region on
chestnuts, and in one case on two specimens of Quercus pedunculata
in Berkeley. The stem mothers appear during the late spring, in
April and May. Viviparous generations are produced throughout the
summer, the sexuales occurring in October and November.

]6. Myzocallis coryli (Goetze)

Figures 43, 44, 53, 54

Goetze, Ent. Beitrage, vol. 2, p. 311, 1778. Aphis (orig desc.).
Clarke, Can. Ent., vol. 35, p. 249, 1903. Callipterus (list).
Davis, Jour. Econ. Ent., vol. 3, p. 417, 1910. Callipterus (desc.).
Essig, Pom. Jour. Ent., vol. 4, p. 762, 1912 (list).

Records. — Corylus sp., Berkeley (Clarke) ; Corylus rostrata, San Francisco Bay
region (Davidson) ; C. rostrata var. calif ornica, C. maxima, Berkeley, August,
1914, June to July, 1915.

In the San Francisco Bay region this species is quite common on
alder. During the seasons of 1914 and 1915 the author observed it
to be very abundant on species of alder on the University of California
campus. He has never found it in the south, however.

17. Myzocallis discolor (Monell)

Figures 262, 263

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 30, 1879. Callipterus (orig.

desc.).
Williams, Univ. Neb. Studies, vol. 10, p. 115, 1910. Callipterus (desc.).

Record. — Quercus macrocarpa, Sacramento, October, 1916 (Davidson).

The author received specimens of this species from Davidson,
which were found in October, 1916, on Quercus macrocarpa in Sacra-
mento. The determination was made by Davis. Below are a few
descriptive notes to supplement Williams' description listed above.

Alate viviparous female. — Antennae about as long as body, III
the longest segment, followed by IV, VI, and V. VI spur is slightly
longer than the base. The antennae are rather slender as compared

26 MISCELLANEOUS STUDIES

with other species of this genus. Primary sensoria are present on
V and VI as usual, and accessory sensoria on VI. There are about
seven secondary sensoria oil III (fig. 262), which are more or less oval
to circular, and located on the basal two-thirds of the segment. The
cornicles, cauda, and anal plate are typical of the genus.

Measurements: Body length 1.28 to 1.37 mm., antenna total 1.41
mm., Ill 0.459 mm., IV 0.306 mm., V 0.264 mm., VI 0.289 mm. (base
0.119 mm., spur 0.17 mm.), cornicles 0.68 mm., wing length 2.074 to
2.414 mm., width 0.68 to 0.833 mm. The two dusky transverse bands
across the fore wings (fig. 263) constitute the most distinguishing
character. The branching of the third discoidal is quite variable.

18. Myzocallis punctatus (Monell)

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 31, 1879. Callipterus (orig.

desc.).

Clarke, Can. Ent., vol. 35, p. 249, 1903. Callipterus Jiyalinus Monell (list).
Essig, Pom. Jour. Ent., vol. 4, p, 762, 1912. M. hyalinus (Monell) (list).
Record. — Quercus imbricata, Berkeley (Clarke).

This is a doubtful species, reported only by Clarke from Quercus
imbricata in Berkeley. It is the author's opinion that this is the same
species listed by Davidson as M. quercus (Kalt.).

19. Myzocallis maureri Swain

Figures 55, 56, 266
Swain, Trans. Am. Ent. Soc., vol. 44, p. 4, 1918 (orig. desc.).

Records.— Quercus agrifolia, Berkeley (Swain) ; Quercus Icelloggii, Julian, San
Diego County (Swain).

This species has been taken in Berkeley and in the Cuyamaca
Mountains of San Diego County by the author. Essig has also taken
it in Berkeley. It is never abundant, but the author has observed it
several times and in several places in the localities mentioned.

20. Myzocallis pasaniae Dvdn.

Figure 57
Davidson, Jour. Econ. Ent., vol. 8, p. 424, 1915 (orig. desc.).

Eecords. — Pasania densiflora, Stevens Creek Canyon, Santa Clara County
(Davidson), Berkeley, February, 1915 (Essig).

This is a species found occasionally on tanbark oak in the San
Francisco Bay region. The author has never taken it but has speci-
mens from Davidson and Essig.

A SYNOPSIS OF THE APHIDIDAE 27

21. Myzocallis quercus (Kalt.)

Figures 31, 32, 58

Kaltenbach, Monog. d. Pflanzenlause, p. 98, 1843. Aphis (orig. dese.).
Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909. Callipterus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911. Callipterus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 762, 1912 (list).
Davidson, Jour. Econ. Ent., vol. 7, p. 130, 1914 (desc.).

Records. — Quercus agrifolia; Stanford University, San Jose, Penryn, Placer
County (Davidson) ; Q. lobata, Santa Clara County (Davidson) ; Berkeley, 1915
(Essig); Q. pedunculata, Berkeley, August, 1914; Q. douglasii, Stanford Univer-
sity, November, 1910, April, 1911 (Morrison); Q. robur, Oakland (Davidson).

This is a variable species more or less common in the San Fran-
cisco Bay region and in the Sacramento Valley on various species of
oaks. When he first reported it Davidson was doubtful of its identity.
Later, however, it was identified by Peter Van der Goot5 as this
species.

22. Myzocallis ulmifolii (Monell)

Figure 59

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 29, 1879. Callipterus (orig.

desc.).

Davidson, Jour. Econ. Ent., vol. 2, p. 301, 1909. Callipterus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Callipterus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 762, 1912 (list).

Records. — Ulmus sp., Stanford University (Davidson), "Ulmus americana, Wal-
nut Creek, October, 1913 (Davidson).

Davidson reports this as common on elms in the San Francisco
Bay region. However, the author has never collected it. The follow-
ing brief descriptive notes are from an alate viviparous female, taken
in Walnut Creek by Davidson. The most distinguishing character is
the presence of a pair of small but prominent tubercles on the mid-
dorsum of the first and second abdominal segments. The usual
primary and accessory sensoria are present on V and VI. Secondary
sensoria (fig. 59) are present on the basal one-half to two-thirds of
III. These are transversely linear or oval, and number about six.
The cornicles are very short, being fully as broad at the apex as long.
Cauda and anal plate normal. Wings normal, radial vein indistinct,
first discoidal curving toward base of wing. Body length 1.836 mm.,
width of thorax 0.578 mm., antennae total 1.309 to 1.326 mm., Ill

s In 1917 George Shhrji (Ent. News, vol. 27, February, 1917) described three
species, M. essiggi n.sp., M. woodworthi n.sp., and M. hyalinus (Monell), all of
which are undoubtedly but varieties of this species, M. quercus (Kalt.).

28 MISCELLANEOUS STUDIES

0.442 mm., IV 0.255 to 0.272 mm., V 0.221 to 0.2465 mm., VI 0.255 mm.
(base 0.136 mm., spur 0.119 mm.), cornicles height 0.034 mm., diam-
eter at apex 0.034 mm., wing length 1.581 to 1.768 mm., width 0.663
to 0.68 mm., expansion 3.825 mm.

8. Genus Chromaphis Walker
Walker, The Zoologist, p. 2001, 1870. Type Lachnus juglandicola Kalt.

23. Chromaphis juglandicola (Kalt.)

Figures 34, 35

Kaltenbach, Monog. d. Pflanzenlause, p. 151, 1843. Lachnus (orig. desc.).
Essig, Pom. Jour. Ent., vol. 1, p. 51, 1909. Callipterus (desc. vivi.).
Essig, Pom. Jour. Ent., vol. 4, p. 763, 1912 (list).
Davidson, U. S. Dept. Agr., Bull. 100, pp. 2-19, 1914 (desc. all forms).

Records. — Juglans regia; San Francisco Bay region, southern California.

This walnut aphis is the most abundant and injurious of the
species attacking walnut in California. It is more or less abundant
throughout the San Francisco Bay region, while in southern Cali-
fornia during certain seasons it is an important pest. Davidson
(1914) has described all the forms and studied the life history care-
fully, so but little comment is necessary. In 1915 the author
observed the young stem mothers on March 22 in Sunnyvale, Santa
Clara County. Three weeks later the second generation was well
advanced. From the first of May on, in 1916, the viviparae were
abundant on walnuts throughout San Diego County, from nursery
stock in San Diego to a few cultivated trees at Santa Ysabel (altitude
3000 feet). From the middle of October until well into December,
1916, the sexuales were found throughout Los Angeles and Riverside
counties.

9. Genus Callipterus Koch
Koch, Die Pflanzenlause, p. 208, 1855. Type Aphis juglandte Kalt.

The two members of this genus in California have been considered
heretofore as species of Monellia Oestlund (genus 10), but according
to Davis6 they can not be so considered for in Monettia the wings are
laid flat on the abdomen when at rest. This is found only in M&nellia
caryella (Fitch). Incidentally it may be remarked that the species
known by that name in California does not have that habit, so
should really be placed in this genus, Callipterus Koch. However, as
it is identical with eastern specimens, except for this habit, the author

6 Essig, E. O., Beneficial and Injurious Insects of California, Mon. Bull. Cal.
Comm. Hort., vol. 4, p. 83, 1915.

A SYNOPSIS OF THE APHIDIDAE 29

has thought best to retain it in Monellis, at least for the time being.

KEY TO CALIFORNIA SPECIES

1. VI spur about equal to or slightly longer than VI base. Tibiae mostly pale.

caryae Monell

— VI spur shorter than VI base. Tibiae entirely dark. Considerably larger than
preceding species calif ornicus (Essig)

24. Callipterus californicus (Essig)

Figures 63, 64

Essig, Pom. Jour. Ent., vol. 4, p. 767, 1912. Monellia (orig. desc.).
Davidson, U. S. Dept. Agr., Bull. 100, p. 34, 1914. Monellia (list, key to

walnut aphids).
Records. — Juglans californica (California black walnut) ; Santa Paula.

In 1912 Essig described this species from specimens taken near
Santa Paula in July, 1911. No other definite collections are known
to the writer, although Essig reports it as more or less abundant on
the California black walnut throughout the southern part of the state.
Davidson has not found it in the San Francisco Bay region, nor has
the author ever observed it, either in the bay region or in southern
California.

25. Callipterus caryae Monell

Figures 65, 66

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 31, 1879 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 249, 1903 (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 301, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).
Essig, Pom. Jour. Ent, vol. 4, p. 764, 1912. Monellia (list).
Davidson, U. S. Dept. Agr., Bull. 100, pp. 19-26, 1914. Monellia (dese.
all forms).

Records. — Juglans regia, J. californica; Berkeley, Stanford University, San
Jose, San Francisco Bay region.

This species is more or less common in the San Francisco Bay
region on walnuts. Davidson has described all the forms and noted
its life history. The author has not taken the species.

10. Genus Monellia Ostluml

Oestlund, Minn. Geol. Nat. Hist. Surv., Bull. 4, p. 44, 1887. Type Aphis
caryella Fitch.

This genus, as described by Oestlund, differs from Callipterus par-
ticularly in the position of the wings when the insects are at rest. In
Callipterus they are held roof-like over the body as is usual in aphids.
but in Monellia they are laid flat on the abdomen. It includes but
the one species, M. caryella (Fitch).

30

MISCELLANEOUS STUDIES

26. Monellia caryella (Fitch)

Figures 25, 67, 68

Fitch, Insects N. Y., vol. 1, p. 163, 1855. Aphis (orig. desc. apt. vivL).
Fitch, Ins. N. Y., vol. 3, p. 448, 1856. Callipterus (first desc. ala. vivi.).
Davidson, Jour. Econ. Ent., vol. 7, p. 132, 1914 (list).
Davidson, U. S. Dept. Agr., Bull. 100, pp. 26-34, 1914 (desc. all forms).

Records. — Juglans calif ornica, J. Nigra, J. regia; San Jose, Walnut Creek
(Davidson) ; Stanford University, May to June, 1915.

A more or less common species on both the native black walnut,
and the cultivated walnut in the San Francisco Bay region. This
species, while very similar to the preceding species, is probably the
more common of the two. The following table of differences is taken
from Davidson:7

Form
Alate viviparous female

Callipterus caryae Monell Monellia caryella (Fitch)

Pupa of viviparous

female
Oviparous female

Antennal joint III very
slightly thickened bas-
ally.

Sensoria on antennal

joint III occupying

basal half or two-
thirds.

Antennal joint VI and

its spur or filament

subequal, or VI less
than spur.

Dusky knee spots often
present.

Four longitudinal rows
of capitate spines.

Smaller than viviparous
female.

Four longitudinal rows
of cipitate spines.

Antennal joint III quite

noticeably thickened
* for its basal half.

Sensoria on antennal
joint III occupying
basal third.

Antennal joint VI one-
third as long again as
its spur or filament.

Dusky knee spots absent.

Six longitudinal rows of
capitate spines.

Larger than viviparous
female.

Six longitudinal rows of
capitate spines.

This species is distinct from the preceding and according to Mor-
rison, who has examined eastern species, is structurally identical
except in the matter of the wings. He writes as follows :

i Davidson, W. M., Walnut aphides in California, II. S. Dept. Agr., Bull. 100,
p. 28, 1914.

A SYNOPSIS OF THE APHIDIDAE 31

I made a very careful study of specimens from California, sent me by David-
son, and of specimens collected both in Indiana and New York (type locality).
I was unable to find any structural differences that would definitely separate the
two lots of specimens, with the exception of the position of the wings. These are
laid flat when at rest in the eastern specimens, but are not so in the Californian
specimens, according to Davidson. In spite of this apparent agreement, I feel
that the two must be distinct.

If this is the case, that the wings are not laid flat at rest, this
species must belong to the genus Callipterus, and therefore cannot be
Monellia caryella (Fitch). However, the author has not had an
opportunity to study this carefully, so leaves it as it is, calling this
California species Monellia caryella (Fitch).

Because of the fact that all the species of aphids on walnut are
so closely related, and so very similar in structure, a key to separate
them, one from another, is given here. This key is adapted from
Davidson.8

1. Cornicles quite evident, about as long as wide.

Chromaphis juglandicola (Kalt.)

— Cornicles barely perceptible, considerably wider than long * 2

2. Tibiae of alate viviparae entirely dusky Callipterus calif ornicus (Essig)

— Tibiae of alate viviparae mostly pale 3

3. VI spur longer than VI base. Oviparous females with four longitudinal rows

of capitate hairs Callipterus caryae Monell

— VI spur shorter than VI base. Oviparous females with six longitudinal rows

of capitate hairs Monellia caryella (Fitch)

11. Genus Callipterinella Van der Goot

Van der Goot, Zur Systematic der Aphiden, 1913. Type Aphis (Callipterus)
betularius Kaltenbach.

27. Callipterinella ammlata (Koch)

Koch, Die Pflanzenlause, p. 1855. Chaitophorus (orig. desc.).
Gillette, Jour. Econ. Ent., vol. 3, p. 367, 1910, Chaitophorus betulae (Buck-
ton) (list).
Davidson, Jour. Econ. Ent., vol. 10, p. 292, 1917 (desc.).

Eecords. — Betula alba; Oakland, Walnut Creek (Davidson).

This species has been reported by Davidson as infesting the leaves
and shoots of the white birch in the San Francisco Bay region. It is
unknown to the author.
s Ibid., p. 35.

32 MISCELLANEOUS STUDIES

Tribe Chaitophorini Wilson (Lachnidea Mordw. and
Chaitopheri Mordw.)

Wilson, Can. Ent., vol. 42, pp. 385-387, 1910.

This tribe as considered by Wilson contains the following genera :
Arctaphis, Chaitophorus, Symydobius, Thoniasia, and Sipha. The
author has followed Wilson's classification, having added, however,
two genera described later by Essig: viz., Micrella and Fullawaya.
Essig's genus Eichochaitophorus is a synonym of Arctaphis Walker
(see discussion under no. 27). Mordwilko's groups Lachnoidea and
Chaitophori are both included in this one tribe. In the former, Mord-
wilko includes Symydobius and Pterochlorus, and in the latter,
Cladobius, Melanoxanthus, and Chaitophorus. Both Cladobius and
Melanoxanthus are included in this paper in the tribe Pterocommini,
being synonyms of the genus Pterocomnia Buckton. Following is a
description of the tribe Chaitophorini as given by Wilson (op. cit.) :

Antennae, except in Sipha, always six-segmented; in Sipha there are but five.
Length variable; antennal tubercles wanting; antennae, legs, and body covered
with hair-like bristles. Fore wings with two oblique veins and cubitus always
twice forked; hind pair with two cross veins. Nectaries (cornicles) variable in
length and size, but never longer than one-tenth the length of the body. The
genera in this tribe are somewhat similar to those in the tribe Callipterini, but
are easily distinguished by the shorter and heavier antennae and legs, as well as
by the finer and more hair-like bristles.

The following key to the Californian genera has been adapted from
Wilson and Essig :

1. Spur of sixth antennal segment at least three times as long as the segment 2

— Spur not three times as long as the segment. Cauda broadly rounded and

without knobbed tip 4

2. Spur more than five times as long as the segment; cornicles longer than the

base of the sixth segment Chaitophorus Koch

— Spur of sixth segment not more than five times as long as the segment; corn-

icles not longer than the base of the sixth segment 3

3. Cauda a knob on a quadrangular base (fig. 69). Spur about five times as

long as sixth segment Arctaphis Walker

— Cauda tapering to a blunt tip which is usually straight across, not being

rounded or constricted at the base (fig. 70). Spur but slightly more than
three times as long as the sixth segment _ Micrella Essig

4. Spur of sixth segment shorter or scarcely longer than the segment; antennae

nearly as long as the body Symydobius Mordwilko

— Spur considerably longer than sixth segment; antennae about one-half the

length of the body _ 5

5. Cornicles absent ; body with lateral tubercles Fullawaya Essig

— Cornicles present ; lateral body tubercles wanting „ Thomasia Wilson

A SYNOPSIS OF THE APHID1DAE 33

Genus Chaitophorus Koch

Koch, Die Pflanzenlause, p. 1, 1854. Type Aphis aceris Linn.

There are at present no species of this genus in California; most
of the species hitherto placed in it are now considered as belonging
to the genus Thomasia Wilson.

12. Genus Arctaphis Walker
Walker, The Zoologist, p. 2000, 1870. Type aphis populi Linn.

This genus as defined by Wilson is represented in California by
two species: A. viminalis (Monell) and A. populifolii (Essig). The
latter was placed by Essig in a new genus, Eichochaitophorus, but
there is not enough difference between these to warrant a new genus.

KEY TO CALIFORNIA SPECIES

1. Wings hyaline. Three-nine large sensoria on third antennal segment (fig. 71).

IV half as long against as V populifolii (Essig)

— Wings subhyaline. About ten rather small sensoria on III. IV but very
little longer than V viminalis (Monell)

28. Arctaphis populifolii (Essig)

Figures 69, 71

Essig, Pom. Jour. Ent., vol. 4, p. 722, 1912. Eichochaitophorus (orig.

desc.).

Davidson, Jour. Econ. Ent., vol. 3, p. 375, 1910. Chaitophorus populifoliae
(Fitch) (desc. male).
Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911. Chaitophorus populifoliae

(Fitch) (list).

Records. — Populus trichocarpa, Santa Paula (Essig), Berkeley, September,
1915; Populus fremontii, Stanford University and Penryn, Placer County (David-
son) ; Menlo Park, San Mateo County, October, 1914 (Morrison) ; Berkeley, Sep-
tember, 1915; El Cajon, San Diego County, June, 1916; Kiverside, October, 1916.

In 1912 Essig described this species from specimens taken on
Populus trichocarpa at Santa Paula, and placed it in a new genus,
Eichochaitophorus. He separated this genus from Arctaphis for the
following reasons :

According to Wilson the cauda [in Arctaphis] is a knob on a quadrangular
base. The anal plate is broadly rounded. In the new genus [Eichochaitophorus]
the style has a distinct neck and is situated on a very distinct conical base. The
anal plate is deeply notched in the middle so as to make it somewhat forked as
in the genus Callipterus.

34 MISCELLANEOUS STUDIES

Although the anal plate is somewhat notched, there is scarcely
difference enough to warrant the forming of a new genus. In fact,
in many specimens one cannot tell whether or not a notch is present.
As to the cauda, consisting of the tip, a distinct neck, and a distinctly
conical base, this is not greatly different from a cauda consisting of
a knobbed tip on a quadrangular base. The only practical difference
is in the base, being conical in one and quadrangular in the other. In
populifolii (Essig) the base seems to be conical, yet one cannot be
certain unless the specimen is mounted exactly.

This species, A. populifolii (Essig), as stated above, was described
from specimens taken on Popuhis trichocarpa at Santa Paula. In
1910 Davidson found a species on Populus fremonti at Stanford Uni-
versity, and the following year at Penryn, Placer County, which he
listed as Chaitophorus populifoliae (Fitch). A careful study of
specimens from Davidson and the cotypes of Essig 's species convinced
the author that they were identical. Morrison writes that Davidson's
specimens are not C. populifoliae (Fitch), so Essig 's species is distinct.
In September, 1915, the author observed a great number of specimens
of this species on a weeping elm (TJlmus sp.) in Berkeley, which was
in close proximity to some populars. However, none were seen to be
feeding on the elm, all being restless and wandering over the leaves
and branches. In southern California this is often found infesting
the empty galls of Thecabms populimonilis Riley, such having been
observed in San Diego and Riverside counties.

29. Arctaphis viminalis (Monell) ?

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 31, 1879. Callipterus (orig.

desc.).

Clarke, Can. Ent., vol. 35, p. 248, 1903. Chaitophorus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 375, 1910. Chaitophorus (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. Chaitophorus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 716, 1912. Thomasia (list, key to Califor-

nian species of Thomasia).
Patch, Maine Agr. Exp. Sta., Bull. 213, p. 80, 1913. CJiaitophorus (desc.).

Records — Salix spp. ; Watsonville, Santa Cruz County, and Newcastle, Placer
County (Clarke); Penryn, Placer County, and Stanford University (Davidson).

This species has been reported from Placer, Santa Clara, and
Santa Cruz counties on various species of willow. The true Chai-
tophorus viminalis Monell is an Arctaphis, but whether or not the
western species is the same as the eastern is a question. The author
has never seen specimens of either and is therefore unable to make

A SYNOPSIS OF THE APHIDIDAE 35

any further comment. He once thought the western spe*cies was iden-
tical with Thomasia salicicola (Essig), to which Morrison considers
it very closely related, but Davidson assures him the two are distinct.

13. Genus Micrella Essig

Essig, Pom. Jour. Ent., vol. 4, p. 716, 1912. Type M. monella n.sp.

30. Micrella monella Essig

Figures 70, 72

Essig, Pom. Jour. Ent., vol. 4, p. 717, 1912 (orig. desc.).
Becords. — Salix lasiolepis, Oxnard (Essig); S. laevigata, Santa Paula (Essig).

This species was taken twice by Essig, who described it, in 1910
near Oxnard, and in 1911 near Santa Paula. Since then it has never
again been found. The author has had access to cotype specimens in
Essig 's collection.

14. Genus Fullawaya Essig
Essig, Pom. Jour. Ent., vol. 4, p. 735, 1912. Type F. saliciradicis n.sp.

31. Fullawaya saliciradicis Essig

Figure 75

Essig, Pom. Jour. Ent., vol. 4, p. 737, 1912 (orig. desc.).
Eecord. — Salix laevigata, Santa Paula, August, 1911 (Essig).

On the roots of Mallow near Santa Paula, Essig once found a large
number of aphids, the greater part of which were apterae, although
a few alates were present. Unable to identify them with any known
species, or to fit them into any genus, he described them as this species.
Since then they have not been taken. The author has had access to
cotype specimens in Essig 's collection.

15. Genus Thomasia Wilson

Wilson, Can. Ent., vol. 42, p. 386, 1910. Type Chaitophorus populicola Thomas.

This genus is separated from Chaitophorus principally by the
comparative lengths of the antennae and the comparative lengths of
the spur of the sixth antennal segment. In Chaitophorus (type Aphis
aceris Linn.) the antennae are almost as long as the body, and the
spur of the sixth segment is over five times as long as the base. In
this genus the antennae are but about one-half as long as the body,
and VI spur is but slightly longer than VI base.

36 MISCELLANEOUS STUDIES

KEY TO CALIFORNIA SPECIES

1. Wings hyaline 2

- Wings with veins clouded (fig. 275) populicola (Thomas)

2. Ill longer than VI (including spur) negundinis (Thomas)

— Ill not longer than VI (including spur) 3

3. IV with secondary sensoria crucis Essig

— IV without secondary sensoria salicicola Essig

32. Thomasia crucis Essig

Figure 76

Essig, Pom. Jour. Ent., vol. 4, p. 742, 1912 (orig. desc.).
Records. — Salix macrostaohya, Santa Paula, August, 1911 (Essig).

Essig once found this species on the leaves of willow near Santa
Paula. Since then it has never again been taken. The author has
had access to cotype specimens in Essig 's collection.

33. Thomasia negundinis (Thomas)

Thomas, 111. Lab. Nat. Hist., Bull. 2, p. 10, 1878. Chaitophorus (orig.

desc.).

Sanborn, Kan. Univ. Sci., Bull. 3, p. 35, 1904. Chaitophorus (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. Chaitophorus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 716, 1912 (list).

Records. — Acer negundo, Stanford University (Davidson) ; Salt Marshes, Palo
Alto, May, 1912 (Morrison).

This species of Thanuisia is quite common on box elder in the
vicinity of Stanford University and Palo Alto. The author has never
taken specimens, nor had access to any. Morrison writes that

*

although he has never had access to eastern specimens of T. negundinis
(Thos.) for comparison he is not able to convince himself that the
western species is negundinis. The author is unable to form any
opinion at present, having never seen specimens, hence lists the species
as Davidson has done.

34. Thomasia populicola (Thomas)

Figures 77, 275

Thomas, 111. Lab. Nat. Hist, Bull. 2, p. 10, 1878. Chaitophorus (orig.

desc.).

Essig, Pom. Jour. Ent., vol. 1, p. 98, 1909. Chaitophorus (desc.).
Essig, Pom. Jour. Ent., vol. 4, p. 716, 1912 (list).

Records. — Populus spp., Salix spp., Santa Paula (Essig) ; Eiverside, May, 1917;
Populus sp., Canton, Broadwater County, Montana, July, 1915, R. W. Haegele;
Edna Canon, Boxelder County, Utah, August, 1916, R. W. Doane.

A SYNOPSIS OF THE APHIDIDAE 37

This species has been reported by Essig from Ventura County.
The author has never taken the alates, but has had the opportunity
of examining Essig 's specimens, and specimens from Montana and
Utah taken by Haegele and Doane. It is easily distinguished from
other members of the genus by the broad, dark wing veins.

35. Thomasia salicicola (Essig)

Figure 78

Essig, Pom. Jour. Ent., vol. 3, p. 532, 1911. Chaitophorus (orig. desc.).
Davidson, Jour. Eeon. Ent., vol. 3, p. 375, 1910. ChaitopJiorus nigrae Oest-

lund (?) (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. Chaitophorus nigrae Oest-

lund (?) (list).

Essig, Pom. Jour. Ent., vol. 4, p. 619, 1912 (note).
Essig, Pom. Jour. Ent., vol. 4, p. 716, 1912 (list).

Eecords. — Salix laevigata, Santa Paula (Essig), Salix nigra, Lakeside, San
Diego County, April, 1916; Populus trichocarpa, Santa Paula (Essig); Salix sp.,
Sau Jose, Stanford University, Penryn, Placer County (Davidson), Fillrnore, Ven-
tura County, March, 1911 (Essig).

Essig reported this from Ventura County, and the author has
found it in San Diego County. It was observed to be in large colonies
on the leaves and leaf petioles of the tender growth of willow, in
company with Siphocoryne carpreae (Fabr.). Specimens taken by
Davidson and listed as Chaitopkorus nigrae Oestlund prove to be
identical with this species.

16. Genus Symydobius Mordwilko

Mordwilko, Bap. Lab. Zool. Kap. Imp. Varch. Univ., 1895. Type Aphis
oblonga Heyden.

KEY TO CALIFORNIA SPECIES

1. Anal plate half -moon-shaped 2

— Anal plate bilobed (fig. 271) ; cornicles pale, appearing white in life; antennae

with about six to eight secondary sensoria on III, and one or two on IV
(fig. 272 ) chrysolepis Swain

2. Spur of VI but a short thumb-like projection ; sensoria on III numbering about

six to ten; none on IV agrifoliae Essig

— Spur of VI longer, being equal to or longer than base of VI ; fifteen to twenty

sensoria on III, one or two on IV (figs. 73, 74) 3

3. Antennae for the most part dark, being dark brown or black; spur and base

of VI equal, lateral abdominal tubercles present in apterae.

macrostachyae Essig

— Antennate for the most part pale, being light brown or amber; spur of

VI usually slightly longer than base; apterae without lateral abdominal
tubercles salicicorticis Essig

38 MISCELLANEOUS STUDIES

36. Symydobius agrifoliae Essig

Essig, Univ. Calif. Publ. Entom., vol. 1, pp. 311-317, 1917 (orig. desc.).
Eecords. — Quercus agrifolia; Santa Paula (Essig).

This interesting aphid was taken in Ventura County on live oak
during 1911. It differs from other members of this genus in the
extremely short spur of the sixth antennal segment. The coloration
is very similar to that of the next species, but the length of VI spur
and the fact that the anal plate is not bilobed serves to distinguish it.

37. Symydobius chrysolepis Swain

Figures 269 to 274

Swain, Trans. Am. Ent. Soc., vol. 44, p. 6, 1918 (orig. desc.).
Eecords. — Quercus chrysolepis; Alpine, San Diego County (Swain).

This is a medium sized, brownish colored aphid found in 1916
infesting the terminal twigs and leaf petioles of maul oak in San
Diego County. Its pale white cornicles are very conspicuous, and
serve as a distinguishing character. The anal plate is bilobed, a char-
acter not found in other members of the genus, and one which may
be sufficient for the separation of the species (and 8. albisiphus Davis,
in which the anal plate is also bilobed) from Symydobius into a new
genus. However, the author believes it best to retain them in this
genus at present. The apterous females were found to be heavily
parasitized by the chalcid fly, Closterocerus utahensis Crawford var.
calif ornicus Girault.

38. Symydobius macrostachyae Essig

Figure 73

Essig, Pom. Jour. Ent., vol. 4, p. 727, 1912 (orig desc.).
Eecords. — Salix macrostachya ; Santa Paula (Essig), Fresno, June, 1915.

Twice has this species been taken, once by Essig near Santa Paula
and once by the author along the San Joaquin River near Fresno.
It is found in fairly large colonies on the younger stems of willow.
These colonies consist for the most part of apterae, only a very few
alates being present.

A SYNOPSIS OF THE APHIDIDAE 39

39. Symydobius salicicorticis Essig

Figure 74

Essig, Pom. Jour. Ent., vol. 4, p. 731, 1912 (orig. dese.).
Record. — Salix laevigata; Santa Paula (Essig).

Together with specimens of Fullauwya saliciradicis Essig, this was
taken on willow along the Santa Clara River near Santa Paula in
August, 1911. The colonies are found on the bark near the surface
of the ground either just above or just below it. Essig reports that it
is preyed upon quite extensively by the larvae of an undetermined
species of syrphus fly. The author has had access to cotype specimens
in Essig 's collection.

Group Lachnina Passerini

Passerini, Gli Afidi, 1860.

In this group there are included two tribes, Lachnini Del Guercio
and Pterocommini Wilson, following Wilson. Mordwilko places but
the one tribe Lachnini in this group, including the genus Pterocomma
Buckton in the tribe Chaitophori. However, to the author the group-
ing followed here seems more natural. The following key is adapted
in part from Borner (Sorauer, Pflanzenkrankheiten, vol. 3, p. 665,
1913) :

Sixth antennal segment with a short, thick (thumb-like) projection. Cornicles
conical (fig. 91) or wart-like. Empodial hair short, and oftentimes indistin-
guishable (fig. 79) Tribe Lachnini

Sixth antennal segment with a slender projection (VI spur) which is about as
long as the segment (VI base). Cornicles cylindrical or clavate (figs. 81, 82).
Empodial hair practically as long as the claws (fig. 80). Tribe Pterocommini

Tribe Pterocommini Wilson

Wilson, Ann. Ent. Soc. Am., vol. 8, pp. 347-358, 1915.

This tribe, as considered by Wilson, contains but the one genus,
Pterocomma Buckton. In a former paper (Can. Ent., vol. 43, p. 384,
1910) he recognized two genera : the one, Melanoxantherium Schoute-
den, in which the cornicles were swollen or vasiform, and the other,

40 MISCELLANEOUS STUDIES

Pterocwnnia Buckton, in which the cornicles were cylindrical. He
states in his later paper : "... after having further studied the group
I am of the opinion that such a divsion is illogical, and if a division
is necessary each species should form a different genus. It, therefore,
seems more practical to confine all the species to a single genus. ' ' The
characters of this tribe and genus are as follows :

Antennae with six segments and reaching near the base of the abdomen.
Wings normally with venation as in Aphis. Nectaries [cornicles] short, but
clavate. Cauda short and broadly rounded at the tip as in Lachni?ii. Entire
body, antennae, and legs covered with long hairs as in Lachnini. As has already
been pointed out by Oestlund, this group appears intermediate between the Clw.Uo-
phorini and the Lachnini. Their habits and actions being in different ways similar
to both.

17. Genus Pterocomma Buckton

Buckton, Monog. Brit. Aphides, vol. 2, p. 143, 1879. Type P. pilosa
Buckton.

KEY TO CALIFORNIA SPECIES

1. Cornicles abruptly constricted at distal end, and without a distinct flange

(fig. 81), the diameter of the opening being less than the diameter of the
smallest part of the cornicle. Wing veins broad and shaded.

flocculosa (Weed)

— Cornicles not so abruptly constricted and with a distinct flange. Wing

veins normal 2

2. Cornicles about twice as long as their greatest diameter ....smithiae (Monell)

— Cornicles considerably longer than greatest diameter, and longer than hind

tarsus populifoliae (Fitch)

40. Pterocomma flocculosa (Weed)

Figure 81

Weed, Insect Life, vol. 3, p. 291, 1891. Melanoxanlhus (orig. desc.).
Wilson, Ann. Ent. Soc. Am., vol. 8, p. 350, 1915 (desc.).

Records. — Salix sp., Berkeley, March, 1915; 1916 (Essig).

In his paper on Pterocomma Wilson states that this species does
not occur on the Pacific Coast. However, in March, 1915, the author
found it rather abundantly on willow on the campus of the University
of California in Berkeley. During the 1916 season Essig observed it
to be quite common in Berkeley. The species is easily recognized in
life by the white cottony flocculence covering the colonies on the bark.

A SYNOPSIS OF THE APHIDIDAE 41

41. Pterocomma populifoliae (Fitch)

Figures 82. 83

Fitch, Cat. Homop. N. Y., p. 66, 1851. Aphis (orig. dese.).

Davidson, Jour. Econ. Ent., vol. 2, p. 300, 1909. Cladobius rufulus n.sp.

(desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 375, 1910. Cladobius rufulus Dvdn.

(list).
Essig, Pom. Jour. Ent., vol. 4, p. 786, 1912. Melanoxantherium rufulum

(Dvdn.) (desc.).
Wilson, Ann. Ent. Soe. Am., vol. 8, p. 353, 1915. Pterocomma populea

(Kalt.) (desc.).
Baker, Can. Ent., vol. 48, pp. 280-282, 1916 (desc.).

Records. — Salix sp. ; Stanford University (Davidson); Santa Paula (Essig);
Walnut Creek, March, 1915 (Davidson) ; Grossmont, San Diego County, March,
1916; Lakeside, San Diego County, April, 1916; Stanford University, May, 1912
(Morrison); Populus sp. ; Stanford University (Davidson); Palo Alto, March,
1915; Populus caroliniana, Banning, Riverside County, April, 1917.

This is a widely distributed species in California on various species
of poplars and willows. Davidson first found it in 1909, describing
it as a new species. In 1915 Wilson stated that it was synonymous
with P. populea (Kalt.), but specimens sent him by the author he
determined as P. bicolor (Oestlund). According to his paper the
cornicles of populea (Kalt.) are about equal in length to the hind
tarsi. Californian specimens have the cornicles considerably longer
than the hind tarsi, but not twice as long as he states they are in
bicolor (Oestlund). His figures of the antennae show that in populea
VI base and spur are subequal, and in bicolor the spur is considerably
longer than the base. The latter is true for the Californian species.
His color notes of populea fit the Californian species very well. Baker
identified Aphis populifoliae Fitch as a Pterocomma and places
rufulus (Davidson) as a synonym. From a study of specimens taken
in Santa Paula, Grossmont, Lakeside, Stanford University, and Wal-
nut Creek, the author finds that Baker's description of populifoliae fits
this species very well. Below are the measurements in microns of
four alate specimens, together with the measurements of cornicles,
antennae, and hind tarsi of one from Lakeside. (This was preserved
for several months in alcohol before being mounted for study, and had
shrunk considerably.)

An examination of the following table shows that in the California
specimens the cornicles are always considerably longer than the hind
tarsi, but never twice as long, and that the spur of six is always longer
than the base, except in one case. This specimen is considerably

42

MISCELLANEOUS STUDIES

.2 "5,

co co in Tf<

a «B 2

s w* ^

t~t-<M.-iaO<Ni-li-l
OOOOrH(MCOt^(M<M
rHrHCM(M<N<M<M<M

3 -s

O O T)< rj< OS tO
CO CO CO CO CM CO

M CO CO 00 CO

CM co co co

rH rH rH rH CM <M (M

Tf( .-I t- O

OOoOOi— I

5— 03 05

-*-» C3 O

!2 o Ir>

O •<+! (M O

O O O rH

CM CM (M <M

O
§

QQ

10

i? S b> '§
-^"S

m .a ^3

•fi

$ O ®

S 05 .S ic

«! -3

OQ

ei

>
S
rP
«

CS «H *? «H

rs 2 s

83 -_

cj c

'Jl

S ^ e«

A SYNOPSIS OF THE APHIDIDAE 43

smaller than the others and has many more secondary sensoria, being
a male. From this evidence this species is the same as Baker lists as
P. populifoliae (Fitch) and should be so considered. The author has
reared a number of specimens of a species of Aphidiiw from material
obtained near Stanford University in May, 1915.

42. Pterocomma smithiae (Monell)

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 32, 1879. Chaitophorus (orig.

desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 300, 1909. Cladobius salicti (Harris)

(list).
Davidson, Jour. Eeon. Ent., vol. 3, p. 375, 1910. Cladobius salicti (Harris)

(list).
Essig, Pom. Jour. Ent., vol. 4, p. 786, 1912. Melanoxantherium salicti

(Harris) (list).
Wilson, Ann. Ent. Soc. Am., vol. 8, p. 355, 1915 (desc.).

Records. — Salix spp., Stanford University (Davidson, Morrison).

Both Davidson and Morrison have taken this species in the vicinity
of Stanford University on various species of willow. According to
Wilson, with whom Morrison and Baker agree, this is P. smithiae
(Monell), the salicti of Harris being synonymous. The sexuales were
observed by Davidson in October, the eggs hatching in January.

Tribe Lachnini Del Guercio

Del Guercio, Eedia, vol. 5, 1908.

This tribe is represented in California by three genera, viz., Essig-
ella Del Guercio, Tuberolachnus Mordwilko, and Lachnus Burmeister,
while there are six genera included in the tribe as it is here considered.
Following is a brief characterization of the tribe adapted from Mord-
wilko :

The body and appendages are very hairy, and usually quite large. The cauda
is absent, the cornicles cupola-shaped, being black or brown in color. Sometimes
they are reduced to mere pores or not fully developed [Lachnus taxifolia Swain].
The antennae in general are not longer than the head and thorax, six- jointed
[except in Essigella Del Guercio], with the spur of the sixth segment very short,
not being as long as the segment itself. The beak is almost always elongated,
generally reaching to or beyond the middle of the abdomen. All this group possess
the anatomical peculiarity that the narrowed hind end of the stomach is covered
with the intestine. The stigma of the fore wing is elongate linear [in Longi-
stigma Wilson it reached past the tip of the wing (fig. 89)]. The cubitus is twice-
branched.

44 MISCELLANEOUS STUDIES

All the California species with the exception of Tuberolachnus vimi-
nalis (Fonsc.), which lives on willow, are found on conifers — Firms
sp., Pseudotsuga sp., or Picea, sp.

Following is a key to the genera, adapted from Del Guercio, Wil-
son and Essig. In this key are included not only the California
genera but the other three as well, in that an understanding of the
characters is thus made easier.

1. Antennae six-segmented 2

— Antennae five-segmented (fig. 83) Essigella Del Guercio

2. Stigma exceptionally long, reaching beyond the tip of the wing (fig. 84).

Longistigma (Wilson)

— Stigma not exceptionally long, not reaching beyond the tip of the wing

(fig. 85 ) 3

3. First joint of the hind tarsus much shorter than half the second (fig. 86) .... 4

— First joint of the hind tarsus equal to or slightly longer than half the second

(fig. 87) -. Eulachnus Del Guercio

4. Abdomen with horn-like tubercle on median dorsum between the cornicles,

(Sometimes this cannot be made out in specimens mounted in balsam, but
it is always readily discernible in fresh or alcoholic material).

Tuberolachmus Mordwilko

— Abdomen without horn-like tubercle 5

5. Bases of first and second discoidal close together; third discoidal often very

faint; wings slightly if ever clouded (fig. 85) Lachnus Burmeister

— Bases of first and second discoidals not so close together as in Lachnus Burm. ;

third discoidal plain; wings often darkly clouded Pterochlorus Rondani

18. Genus Essigella Del Guercio

Del Guercio, Eev. di patal. veg., vol. 3, p. 328, 1909. Type Lachnus cali-
fornicus Essig.

43. Essigella calif ornica (Essig)

Figures 3, 5, 83

Essig, Pom. Jour. Ent., vol. 1, p. 1, 1909. Lachnus (orig. desc.).

Del Guercio, Pom. Jour. Ent., vol. 1, p. 73, 1909 (translation by C. F.

Baker of Del Guercio 's paper listed above).
Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 780, 1912 (desc.).

Records. — Pinus radiata; Claremont, Los Angeles County, and Santa Paula
(Essig); Pinus sabiniana, Stanford University, March, 1915; Pinus spp., Stan-
ford University, March and April, 1912 (Morrison) ; Ontario, San Bernardino
County, January, 1917.

This curious little aphid, described by Essig from specimens taken
in Claremont, Los Angeles County, on Pinus radiata, has since been
found in several parts of the state. Wilson has taken it in Oregon
on Pseudotsuga taxifolia, and Patch in Maine on Pinus strobus. It

A SYNOPSIS OF THE APHIDIDAE 45

is a small, slender, long-legged aphid, that clings fast to the pine
needles and is extremely difficult to see. However, if a branch of
pine is struck sharply and with considerable force over a white paper
or cloth, a large number of these aphids will jar off.

19. Genus Tuberolachnus Mordwilko

Mordwilko, Ann. Mus. Zool. d. 1'Acad. Imp. Sci., vol. 13, p. 374, 1908.
Type Aphis viminalis Fonsc.

44. Tuberolachnus viminalis (Fonsc.)

Figure 86
/
Boyer de Fonscolmbe, Ann. Ent. Soc. France, vol. 1O, p. 162, 1841. Aphis

(orig. desc.).

Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909. Lachnus (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. Lachnus (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. Lachnus dentatus Le

Baron (list).
Essig, Pom. Jour. Ent., vol. 4, p. 774 (772), 1912 (list).

Records. — Salix spp., Stanford University and Penryn, Placer County (David-
son); Ventura County (Essig); Stanford University, November, 1914; Berkeley,
July, 1915; Eiverside, July, 1916.

This extremely large aphid, which lives in large colonies on the
branches of various species of willows, is found throughout the San
Francisco Bay region, Sacramento Valley, and southern California,
although it is not at all common. Davidson reports considerable
parasitization by a species of Epherdius, and Essig infection from
some bacterial or fungus disease. The large size and the presence of
a dorsal abdominal tubercle are distinguishing characters.

20. Genus Lachnus Burmeister

Burmeister, Handbuch d. Entomologie, p. 91, 1835. Type Lachnus faciatus,
n.sp.

This is the third largest genus of aphids in regard to the number
of species in California. All the species are to be found on various
conifers, usually feeding through the bark of the branches or trunk.
Characters for distinguishing the species are hard to obtain, and
those used by the author in the following key are of no value except
with specimens of the alate viviparae. This key is not at all adequate,
and is offered here merely as an aid. The author understands that
Wilson is preparing a monograph of this genus, which will undoubt-
edly prove quite valuable.

46 MISCELLANEOUS STUDIES

KEY TO CALIFORNIA SPECIES

1. Beak reaching considerably beyond the third coxa 2

— Beak at most barely reaching to the third coxa 8

2. Beak reaching almost to or. even beyond the tip of the abdomen 3

— Beak not reaching to the tip of the abdomen 4

3. First joint of hind tarsus more than one-third as long as the second joint.

Legs black except the base of the femora and a broad ring near the base
of the tibiae ponderosa Williams

— First joint of hind tarsus scarcely more than one-fourth as long as the second

joint. Legs pale at the base of the femora and tibiae, black at tips.

oregonensis Wilson

4. Body exceptionally Jarge, being over 4 mm. long, usually about 5 mm., and

over 2 mm. wide _ 5

— Body of average size, being from 2.5 mm. to 3 mm. long, and from 0.75 to

1.2 mm. wide 7

5. Third segment of antennae with many sensoria (eight or more), (figs. 88,

89) 6

. — Third joint of antennae with but few or no sensoria, at most with one or two.
First joint of hind tarsus a little less than half as long as the second.
On Pinus sdbiniana sabinianus n.sp.

6. Third joint of antennae with about 8-12 sensoria (fig. 88). Tibiae with a pale

ring near the base. First joint of hind tarsus scarcely more than one-third
the length of the second. On Picea sp vanduzei n.sp.

— Third joint of the antennae with 19-20 sensoria (fig. 89). Tibiae without

pale ring near base. First joint of hind tarsus almost one-half the length
of the second. On Pinus sp. and Abies sp ferrisi Swain

7. Beak not reaching to the middle of the abdomen. Segment three of the

antennae almost as long as the fourth, fifth, and sixth together. Apex
of stigma meeting the margin of the wing in an acute angle, and not
terminated by a distinct vein (fig. 92). On Pseudotsuga taxifolia.

pseudotsugae Wilson

— Beak reaching beyond the middle of the abdomen. Third antennal segment

not nearly so long as the fourth, fifth, and sixth together. Apex of stigma
meeting the wing margin in an obtuse angle, and terminated by a distinct
vein (fig. 93). Apterous viviparous females with a distinctive pattern on
dorsum of abdomen. On Thuya occidentalis tujafilinus (Del Guercio)

8. First joint of hind tarsus longer than one-fourth the second 10

— First joint of hind tarsus less than one-fourth the second 9

9. Third antennal segment without sensoria (fig. 94). Body robust, being of the

usual Lachnus shape. Third discoidal twice-branched, only occasionally
once-branched. On Abies grandis occidentalis Davidson

— Third antennal segment with several irregular sensoria (fig. 95). Body long

and narrow, being somewhat the shape of Essigella calif ornica (Essig).
Third discoidal simple or once-branched. On Pinus sp.

pini-radlatae Davidson

10. Cornicles very poorly developed, seemingly absent in some cases (fig. 103).
Segment three of antennae with five-seven large circular sensoria which are
hardly distinguishable (fig. 106). On Pseudotsuga taxifolia.

taxifolia Swain

— Cornicles normal (fig. 97), being quite conspicuous. Third antennal segment

with two-four clearly defined sensoria (fig. 101). On Picea glehni.

glehnus Essig

A SYNOPSIS OF THE APHIDIDAE 47

45. Lachnus ferrisi Swain

Figures 89, 91

Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909. Lachnus abietis Fitch

(list).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. LacUnus dbietis Fitch

(list).

Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912. Ladhnus abietis Fitch (list).
Swain, Trans. Am. Ent. Soc., vol. 44, p. 9, 1918.

Records. — Abies concolor, Stanford University (Davidson) ; Pinus sp., Stan-
ford University (Swain).

This large lachnid, recently described by the author, has been
found only in the vicinity of Stanford University, in 1909 and 1910
by Davidson on lowland fir, and in 1915 by Ferris on some young
pine trees. Since then it has not been observed.

46. Lachnus glehmis Essig

Figures 96, 97

Essig, Pom. Jour. Ent. Zool., vol. 7, pp. 180-187, 1915 (orig. desc.).
Eecord. — Picea glehni, Sacramento (Essig).

Essig described this species from specimens taken on a Japanese
spruce in Capitol Park, Sacramento, in 1912. At the time it was so

abundant that control measures were deemed necessary. The author

\

has had access to the type specimens in Essig 's collection.

47. Lachnus occidentalis Davidson

Davidson, Jour. Econ. Ent., vol. 2, p. 300, 1909 (orig. dese. apterae).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912 (list).
Wilson, Can. Ent., vol. 44, p. 193, 1912 (desc. all forms).

Records. — Abies grandis, Stanford University (Davidson, Morrison, Ferris
and the author) ; Abies concolor, Corvallis, Oregon (Wilson).

This species is practically always present on a lowland fir tree in
the cactus garden of the Stanford University grounds. Wilson has
found it in the vicinity of Corvallis, Oregon, on white fir. Davidson
states that it is heavily preyed upon by the larvae of Syrphus arciuvtus
and Syrphus opinator.

48 MISCELLANEOUS STUDIES

48. Lachnus oregonensis Wilson

Wilson, Trans. Am. Ent. Soc., vol. 12, p. 103, 1915 (orig. desc.).
Eecord. — Pinus contorta, Oregon and California (Wilson).

There has been no published record of this species from California.
Wilson wrote the author some time ago that he had taken it in this
state, although he gave no definite locality. The author has never seen

specimens.

•

49. Lachnus pini-radiatae Davidson

Figure 95

Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910 (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911 (list).
Essig, Pom. Jour. Ent,, vol. 4, p. 773, 1912 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 785, 1912 (descriptive note).

Becords. — Pinus radiata, Stanford University (Davidson), August, 1914, April,
1915 (author), March, 1916 (K. B. Brown); Pinus ponderosa, Bowman, Placer
County, November, 1911 (H. H. Bowman), Berkeley, March, 1915 (Geo. Shinji) ;
Pinus sabiniana, Penryn, Placer County (Davidson).

This is a fairly small, slender-bodied, long-legged lachnid found
infesting the needles of various pines in the San Francisco Bay region
and in the Sacramento Valley. They are easily recognized on the
needles by the whitish mass of flocculence which covers their bodies.

50. Lachnus ponderosa Williams

Figure 104

Williams, Univ. Neb. Studies, vol. 10, p. 106, 1910 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. 127, 1914 (list).

Eecord. — Pinus ponderosa jeffreyi, Tallac, Eldorado County (Davidson).

Davidson's is the only report of this species in California. The
identification of his specimens was verified by Davis. One specimen
the author saw was quite small, being much smaller than the others
taken by Davidson.

51. Lachnus pseudotsugae Wilson

Figures 92, 98

Wilson, Can. Ent., vol. 44, pp. 159, 302, 1912 (orig. desc.).
Eecord. — Pseudotsuga taxi folia; Oregon, California (Wilson).

A SYNOPSIS OF THE APE I DI DAE 49

Wilson wrote the author some time ago that he had taken this
species in California, although he gave no definite locality or collec-
tion record. The author has had the opportunity to study cotype
specimens.

52. Lachnus sabinianus n.sp.

Eecord. — Pinus sabinian-a, San Francisco (Compere).

In March, 1915, Harold Compere of the California State Insectary
found a small infestation of a species of Lachnus on Digger Pine in
the Golden Gate Park, San Francisco. Since this one collection, the
species has not again been observed. Being unable to identify the
species with any described in America, a description is herewith
appended, the species being named after its host plant, Pinus sabin-
iana. All the specimens, including the types are in the collections
of E. 0. Essig and of the University of California, Berkeley. The
specimens were all mounted in Canadian balsam before color notes
were taken, so those in the following description are only approxi-
mately correct.

Alate viviparous female. — Rich chestnut-amber to dark brown.
Antennal segments I and II, amber; III, yellowish with tips darker;
IV, V, and VI, dark yellow to dusky. Prothorax, chestnut-brown.
Thoracic lobes very dark brown to black. Beak, pale with tips dusky.
Cornicles, black. Cauda and anal plate with distal margins black.
Femora, chestnut-brown with base amber; tibiae, brown with amber
ring near the base; tarsi, amber. Wing veins, grayffi stigma, dusky

Measurements : Body 4.2 mm. long and 1.7 mm. wide at thorax.
Antennae reach to base of abdomen, without secondary sensoria. I,
0.10 mm. ; II, 0.09 mm. ; III, 0.50 mm. ; IV, 0.25 mm. ; V, 0.19 mm. ;
VI, 0.08 mm. ; total, 1.21 mm. Beak reaches to the base of the cor-
nicles. Cornicles medium sized and of the usual Lachnus shape,
being

Apterous viviparous female. — Chestnut-brown in color with black
dorsal spots on abdomen. Antennal segments I and II, dark; III,
dusky yellow with tip dark; IV, V, and VI slightly darker. Beak
reaches to the base of the cornicles. Coxae, black ; femora, black with
basal one-fifth paler; tibiae, black with pale ring near base; tarsi,
black. Cornicles, black and conspicuous. They measure 5.2 mm. in
length and 3.3 mm. in width.

50 MISCELLANEOUS STUDIES

53. Lachnus taxifolia Swain

Figures 99-103
Swain, Trans. Am. Ent. Soc., vol. 44, p. 11, 1918.

Eecords. — Pseudotsuga taxifolia, Sacramento (Essig), Berkeley and San Fran-
cisco (Shinji).

This is a fairly qommon species found in colonies on the branches
and trunks of Douglas fir in the San Francisco Bay and Sacramento
Valley. It is interesting particularly because of the atrophied cor-
nicles.

54. Lachnus tujafilinus (Del Guercio)

Figures 93, 105

Del Guercio, Eedia, vol. 5, p. 287, 1909. Laclmeilla (orig. desc.).

Essig, Pom. Jour. Ent., vol. 3, p. 541, 1911. Lachnus juniperi DeGeer

(desc.).
Essig, Pom. Jour. Ent., vol. 4, p. 773, 1912. Lachnus juniperi DeGeer

(list).
Davidson, Jour. Econ. Ent., vol. 7, p. 127, 1914 (list).

Eecords. — Thuya occidentalis, Claremont, Santa Paula (Essig); Palo Alto,
Walnut Creek (Davidson); Stanford University, March, 1912 (Morrison); San
Diego, March, 1916; Riverside, October, 1916, March, 1917.

This oddly marked Lachnus is more or less common throughout
California wherever arborvitae is cultivated. The apterous females
are the most common, and are easily recognized by the odd markings
on the dorsum of the abdomen (see Essig 's illustrations). Occasion-
ally the alate females are found, Davidson finding some in April,
Morrison and the author in March. The author has observed the
larvae of Coccinella calif ornica feeding on them in Riverside.

55. Lachnus vanduzei n.sp.

Figure 88
Records. — Picea sp., Berkeley, September, 1914 (Essig, E. P., Van Duzee).

In September, 1914, E. P. Van Duzee collected a few specimens
of a large Lachnus on a species of spruce in Strawberry Canyon, near
Berkeley. Later in the same month Essig found specimens on the
same tree. The following fall the author hunted for the species,

A SYNOPSIS OF THE APHIDIDAE 51

but was unable to find any specimens, the tree on which it was first
found having been cut down. In the following description the color
notes are not absolutely accurate, as they were taken from material
mounted in balsam. This species is named after its first collector,
Mr. E. P. Van Duzee, of the University of California. Type speci-
mens are in the collection of the University of California.

Alate viviparous female. — The alate viviparous females are of a dark
muddy color, as near as can be judged from the mounted specimens.
The antennae are : I and II, dusky ; III and IV, pale with apical half
dusky ; V, pale with the apex or apical third dusky ; VI, pale with the
apex and spur dusky. The measurements of the segments are : I,
0.09 mm.; II, 0.07 mm.; Ill, 0.5 mm.; IV, 0.26 mm.; V, 0.27 mm.;
VI, 0.16 mm. The sensoria are located as follows : III, 10-12 ; IV,
2-3 ; V, 2-3 ; VI, 1. They are large and circular, and quite evenly
distributed in a line on each segment. The beak reaches to the base
of the cauda. The coxae are black, the femora amber on the basal
half and black on the apical, the tibiae are black with an amber ring
near the base, the tarsi are black. The first joint of the hind tarsus
is not one-third the length of the sedond, the first measuring 0.08 mm.,
and the second 0.26 mm. The wings are quite large, with a very
distinct stigma. The costal vein is grayish-brown, the subcostal
brown. The stigma is long and brown, the stigmal vein being pale
brown and slightly curved throughout its entire length. The first
and second discoidals are distinct and pale brown, the second dis-
coidal being slightly curved near the tip. The third discoidal is indis-
tinct and twice-branched, the angles of the branches being very acute.
Apterous viviparous female. — Prevailing color, amber-brown, with
the abdomen mottled gray, brown, and black. The head is brown
with anterior margin amber. The antennae are colored as follows:
I, amber; II, amber; III, amber with tip dusky; IV, amber with tip
dusky ; V, amber with apical two-thirds dusky ; VI, dusky. The beak
reaches to the base of the cauda. The femora are brown with the
bases amber, the tibiae and tarsi brown. The first joint of the hind
tarsus is scarcely more than one-third the length of the second. In
four tarsi measured, the relative lengths of the joints were: 0.07 to
0.23 mm. ; 0.08 to 0.23 mm. ; 0.08 to 0.28 mm. ; and 0.07 to 0.25 mm.
The cornicles are conspicuous and dark, the cauda well rounded and
dark on its posterior edge. The lengths of the antennal segments are :
I, 0.1 mm. ; II, 0.1 mm. ; III, 0.56 to 0.57 mm. ; IV, 0.21 to 0.23 mm. ;
V, 0.22 to 0.28 mm. ; VI, 0.15 to 0.16 mm.

52 MISCELLANEOUS STUDIES

Group Aphidina Wilson

Wilson, Ann. Ent. Soc. Am., vol. 3, p. 314, 1910.

This group as considered by Wilson consists of three tribes:
Trichosiphini, Macrosiphini, and Aphidini. The first of these con-
tains two genera found only in the Asiatic islands, so it will not be
considered in this p'aper. This group contains quite closely related
genera, and in many cases it is quite hard to distinguish between
them. Following is a brief extract from Wilson's paper (cited above) :

In studying closely related genera the development of the external characters
may be placed in five divisions: (1) the antennae and spur; (2) the antennal
tubercles; (3) the development of the nectaries [cornicles]; (4) the development
of the cauda; (5) the development of the wing venation. In a group of insects
as pliable as the present one, any one or two of these characters may be either
under- or over-developed and it is necessary to place the genera according to the
greatest development. Of all the characters which show this variation the wings
show what may be true of all these characters.

The two tribes have been separated from one another on the character
of the antennal tubercles, as Wilson says in the same paper :

The division is made between species with distinct antennal tubercles and
those having none or at the most indistinct tubercles. However, should a certain
species have distinct antennal tubercles with the other characters [of the Macro-
siphini] wanting, then it would have to go into the next tribe [Aphidini].

The keys to the tribes and genera below have been formulated by the
author, following, however, those of Wilson, Van der Goot, and
Mordwilko.

1. Antennal tubercles well formed. Antennae usually as long as or longer than
the body. Apterae often with sensoria on the third antennal segment.
Body never with lateral tubercles on the seventh abdominal segment. Cor-
nicles variable but usually about one-fourth the length of the body or

longer Tribe Macrosiphini

— Antennal tubercles absent or more or less indistinct. Antennae seldom longer
than the body. Apterae seldom with sensoria on the third antennal seg-
ment. Body with lateral tubercles on at least the seventh abdominal seg-
ment Tribe Aphidini

Tribe Macrosiphini Wilson

Wilson, Ann. Ent. Soc. Am., vol. 3, p. 314, 1910.

To a large extent the author has followed Wilson in the placing
of the genera, but in a few cases he has not. This is noticeable in
Toxoptera, which is considered by Wilson as belonging to this tribe,

A SYNOPSIS OF TEE APHIDIDAE 53

while the author feels that it is better associated with the Aphidini,
inasmuch as the antennal tubercles are very small and more or less
indistinct and as the antennae are scarcely as long as the body. Van
der Goot's genus, Myzaphis, has been accepted for the two species,
Myzus rosarum (Walker) and Aphis abietina Walker, and is included
with the Aphidini. The species Aphis nymphaeae Linn., which Wil-
son uses as the type of Rhopalosiphum, has been taken from this genus
and placed in Siphocoryne, chiefly because of the apparent absence
of antennal tubercles and of the presence of distinct tubercles on the
seventh abdominal segment. Therefore Aphis persicae Sulzer takes
the place as type of the genus Rhopalosiphum.

KEY TO CALIFORNIA GENERA

1. Cornicles cylindrical, or at most but very slightly swollen on one side (figs.

122, 152 ) 4

— Cornicles distinctly swollen toward apex, or clavate (figs. 109, 113, 119) 2

2. Antennal tubercles very large and tapering but not gibbous on the inner side;

the bases of the antennae being more or less approximate (fig. 107).

Nectarosiphon Schouteden

— Antennal tubercles distinct, but not large and tapering as above, being more

or less toothed or gibbous on the inner side; the bases of the antennae not
approximate (figs. 108, 111) 3

3. Antennal tubercles short and wedge-shaped, the outer side not evident (fig.

108). Cauda ensiform and of medium size. Antennae at most but slightly
longer than the body Rhopalosiphum Koch

— Antennal tubercles short, but not wedge-shaped (fig. 111). Antennae con-

siderable longer than the body. Cauda very large and long.

Amphorophora Buckton

4. Antennal tubercles large and as long on the outer as on the inner side (fig.

106) 5

— Antennal tubercles with outer side shorter than inner, or not evident (figs. 112,

115, 116) 7

5. Cornicles tapering, longer than cauda which is ensiform (fig. 152). "Wing

venation regular, with third discoidal twice-branched.

Macrosiphum Passerini

— Cornicles and cauda variable. Wing venation irregular and very striking with

veins either wanting or combined, and shaded 6

6. Antennal tubercles with short upper inner angle. Cauda shorter than cornicles

and tapering. Stigmal and third diseoidal veins meet in a broad dark
band, giving the wing the appearance of having a closed triangular cell
(fig. 110) Idiopterus Davis

— Antennal tubercles with small rounded tubercle at the upper inner angle.

Cornicles slightly constricted in the middle and at the tip. Wing venation
variable, but usually the stigmal and third discoidal veins are partly
joined and form a distinct, closed, four-sided cell Pentalonia Coquerel

7. Antennal tubercles and first antennal segment with a strong tooth on the

inner side of each (figs. 115, 116). Cauda short and tapering (fig. 118).
Cornicles cylindrical and tapering slightly with tip outcurved (fig. 117).

Rhorodon Passerini

54 MISCELLANEOUS STUDIES

— Antennal tubercles with a distinct but not prominent blunt projection forming
the inner angle (fig. 112), but the prominent teeth as above are lacking.
Cauda short, tapering, and usually triangular (fig. 121). Cornicles as
above, being cylindrical, with a slight tapering from base to apex, and
often slightly outcurved at tip (fig. 122) Myzus Passerini

21*. Genus Amphorophora Buckton

Buckton, Monog. Brit. Aphides, 1876. Type A. ampullata n.sp.
KEY TO CALIFORNIAN SPECIES9

Cornicles pale, or at most slightly dusky, swollen and vasiform (fig. 113). VI
spur longer than III, the latter with 35-45 sensoria riibi (Kalt.)

Cornicles black, greatly dilated in apical one-half (fig. 161). VI spur shorter
than III, latter with 13-17 sensoria latysiphon Davidson

56. Amphorophora latysiphon Davidson

Figure 161
Davidson, Jour. Econ. Ent., vol. 5, p. 408, 1912 (orig. desc.).

Records. — Vinca major, San Jose (Davidson) ; Courtland, Contra Costa County
(Davidson) ; Stanford University, 1912 (Morrison, Essig). Convolvulus arvensis,
San Jose (Davidson). Solarium tuberosum, Walnut Creek, Contra Costa County,
1915 (Davidson).

This species has been found sparingly in the San Francisco Bay
region on periwinkle, morning-glory, and potato tubers, although it
has never seemed to be common. The author has not collected it, his
only specimens being some taken by Essig on periwinkle near Stan-
ford University. The odd shape of the cornicles is a distinguishing
character.

57. Amphorophora rubi Kalt.

Figures 111, 113, 162

Kaltenbach, Monog. d. Pflanzenlause, p. 23, 1843. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 5, p. 411, 1912 (list).
Shinji, Can. Ent., vol. 49, p. 52, 1917 (list).

Records. — Hubus parviflorus; San Jose (Davidson) : Eubus spp., Walnut Creek,
1915 (Davidson); Berkeley (Shinji).

This species has been taken a few times on thimble-berry in the
San Francisco Bay region. Davidson writes that he has also found
it on blackberry and loganberry in the vicinity of Walnut Creek,

»G. O. Shinji (Can. Ent., vol. 49, p. 51, 1917) described an aphid from Ciculta
virosa var. calif ornica in Berkeley, which he called Amphorophora cicutae n.sp.
The author has never seen specimens, so does not feel that he can recognize this
as a good species. Of some half dozen new (?) species described by Shinji the
author has found none, on examining specimens, that are good species, hence he
cannot recognize this one at present.

A SYNOPSIS OF THE APHIDIDAE 55

Contra Costa County. The author has recently received specimens
from Gillette of an alate viviparous female and apterous oviparous
females taken in the vicinity of Fort Collins, Colorado. Inasmuch as
the descriptions of this species are inadequate and not readily acces-
sible it has been thought best to give here brief descriptions of the
different forms. As no color notes were received with the specimens
they must necessarily be omitted.

Alate viviparous female (from Fort Collins, Colorado). — Antennae
half as long again as the body, dusky, and placed on small but distinct
tubercles. From the mounted material it appears as if III were
dusky, IV, pale with extreme tip dusky ; V, pale with apical one-third
dusky; and VI dusky. VI spur is the longest segment, followed by
III, IV, V, VI base, I, and II. The usual primary and accessory
sensoria are present on VI base, and the primary sensorium on V.
Secondary sensoria are present only on III. These are small, circular,
irregular-sized, and irregularly placed along the whole length of the
segment. The number (35 to 40) is such as to make the segment
appear tuberculate. The beak is quite large and long, reaching to or
slightly beyond the third coxae. The thorax is dusky. The wings
fairly large, and normal. The second branch of the third discoidal
vein arises nearer to the base of the first branch than to the apex of
the wing. Normally the measurements are as follows : From the base
of the second branch of the third discoidal to the tip of the wing is
about 0.8 mm., from the base of the first branch to the base of the
second 0.4 mm., from the apex of the first branch to the apex of the
second 0.29 mm. In one case the base of the second branch was 1.02
mm. from the apex of the wing, and but 0.034 mm. from the base of
the second, while the apices of the two branches were but 0.187 mm.
apart. The legs are long, femora pale with apical one-fourth dusky,
tibiae and tarsi dusky. The abdomen is pale with some slight dorsal
dark markings, these being indistinct in the mounted specimens. The
cornicles are fairly long, clavate on the apical one-half or two-thirds,
dusky throughout, and with the extreme tip reticulated. In length
they are somewhat shorter than III, but longer than IV. The cauda
is pale, short, and triangular, being about equal in length to the
hind tarsi.

Measurements : body length, 1.785 mm. ; antennae total, 2.788 mm. ;
III, 0.68 mm. ; IV, 0.51 mm. ; V, 0.408 to 0.425 mm. ; VI, base, 0.12
mm.; VI, spur, 0.867 to 0.884 mm.; cornicles, 0.578 to 0.646 mm.;
cauda, 0.102 mm.; hind tarsi, 0.102 mm.; wing length, 3.128 mm.;
width, 1.292 mm. ; expansion, 6.8 mm.

56 MISCELLANEOUS STUDIES

Apterous oviparous female (Fort Collins, Colorado). — Pale
throughout, with many small hairs scattered over the body. Most
of these hairs are simple, but some especially on the front of the head
and on the bases of the antennae, are capitate. Antennae slightly
longer than the body, pale, with VI and the apices of the other seg-
ments dusky. \l spur and III are subequal or either one may be
slightly longer than the other. These are followed by IV, V, VI base,
I, and II. The usual primary and accessory sensoria are present on
VI base, and the primary sensorium of V. Secondary sensoria are
present only on III, and number about nine or ten. These are small,
circular, but varying in size, and are arranged in a more or less
even line along the basal one-half to two-thirds of the segment. Beak
pale, with tip dusky, quite large and long, reaching to or beyond the
third coxae. Thorax and legs normal, except the hind tibiae which
are quite long, and furnished with a large number of sensoria.
These sensoria cover practically the whole joint. Cornicle very long
and large, curved outward, pale, with apex dusky, and with distinct
reticulations at the extreme tip. They are markedly larger than in
the alate viviparous females, being considerably longer than the third
antennal segment, and in some cases even half as long again. The
cauda is small, pale, and triangular, although somewhat larger in the
viviparous female.

Measurements : bodj* length, 2.04 mm. ; width of thorax, 0.595 mm. ;
antennae total, 2.446 mm. ; III, 0.646 to 0.697 mm. ; IV, 0.442 to 0.459
mm. ; V, 0.356 to 0.374 mm. ; VI, base, 0.136 mm. ; VI, spur, 0.663 mm. ;
cornicles, 0.918 to 0.952 mm. ; cauda, 0.187 mm. ; hind tarsi, 0.136 mm.

22. Genus Idiopterus Davis
Davis, Ann. Ent. Soc. Am., vol. 2, p. 198, 1909. Type, I. neprelepidis n.sp.

58. Idiopterus nephrelepidis Davis

Figure 110

Davis, Ann. Ent. Soe. Am., vol. 2, p. 198, 1909 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).
Essig, Pom. Jour. Ent, vol. 3, p. 538, 1911 (list).

Records. — Nephrolepis exaltata, Santa Paula (Essig), Palo Alto, April, 1915,
San Diego, March to May, 1916; Riverside, February, 1917: Cyrtonium fulcotum,
Berkeley, March, 1915 (Essig); ferns (unidentified species of house ferns), Stan-
ford University (Davidson, Morrison); Viola sp., Claremont (Essig).

This small black aphid is often found in houses and nurseries, and
occasionally out of doors, on the fronds of various kinds of house

A SYNOPSIS OF THE APHIDIDAE 57

ferns, particularly the Boston fern. Essig has also found it on violets
in the vicinity of Pomona College. The alate females have the wings
beautifully marked with black and white.

23. Genus Macrosiphum Passerini
Passerini, Gli Afidi, 1860. Type Aphis rosae Linn.
KEY TO CALIFORNIA SPECIES
Alate viviparous females

1. Cornicles slightly clavate on one side, somewhat as in Bhopalosiphum.

tulipae (Monell)

— Cornicles not clavate 2

2. Ill as long as V and VI (base and spur) sonchella (Monell)

— Ill not as long comparatively 3

3. Ill pale, IV, V, and VI dusky jasminum (Clarke)

— Not so; if IV, V, and VI are dusky then III is also, except perhaps the base;

or if III is pale throughout then at least the greater part of IV and V are
also pale 4

4. VI (base and spur) shorter than III, but V and VI together are longer than

III baccharadis (Clarke)

— VI (base and spur) not shorter than III 5

5. Secondary sensoria on III, IV, and V. Cornicles not reticulated.

heucherae (Oestlund)

— No secondary sensoria on V 6

6. Secondary sensoria on both III and IV. Cornicles with tips at least reticulated

(fig. 152 ) 7

— No secondary sensoria on IV 10

7. Cornicles and cauda subequal in length, the former being more or less bottle-

shaped sanborni Gillette

— Cornicles longer than cauda 8

8. Cauda light green. Secondary sensoria only occasionally present on IV and

then very small and indistinct rosae (Linn.)

— Cauda dark (brown or black). Seven or more distinct secondary sensoria

on IV 9

9. Body with capitate setae, especially on head and antennae.

artemisiae (Fonsc.)

— Body without capitate setae. Abdomen with dark dorsal markings.

lactucae (Kalt.)

10. Cornicles with at least tips reticulated (fig. 132) 14

— Cornicles with no reticulations (fig. 156) 11

11. Body with fan-shaped setae artemisicola (Williams)

— Body without fan-shaped setae 12

12. Distal two-thirds of cornicles black orthocarpus (Dvdn.)

— Only tip of cornicles black 13

13. Cornicles long and slender. About 18 secondary sensoria in a row on III

(fig. 130) pisi (Kalt.)

— Cornicles shorter and heavier. About 25 to 30 sensoria scattered irregularly

along III (fig. 157) dirhodum (Walker)

14. Cornicles with more than apical one-half reticulate (fig. 149).

ludovicianae (Oestlund)

— Cornicles with less than apical one-half reticulated (fig. 128) 15

58 MISCELLANEOUS STUDIES

15. Cornicles dusky for practically their entire length ............................................ 20

— Cornicles with less than apical one-half dusky .................................................. 16

16. Cornicles considerably longer than III, with apical portion curved outward.

About a dozen, medium-sized sensoria in a straight line along basal two-
thirds of III (fig. 131) .................................................. californicum (Clarke)

— Cornicles not considerably longer than III ........................................................ 17

17. Cornicles and VI spur subequal, the former fairly long, slightly curved

outward and Rightly swollen before the tip (fig. 128) ........ Stanley! Wilson

— Cornicles considerably shorter than VI spur, and not swollen before the tip 18

18. Secondary sensoria in a fairly straight line on III. Body not pulverulent 19

— Body covered with a slight pulverulence. Ill with about 30 fairly large-

sized sensoria, more or less scattered along the entire length (fig. 143).

albifrons Essig

19. Cornicles about half the length of VI spur and considerably shorter than III,

the latter with about 20 to 30 secondary sensoria .................. pteridis Wilson

— Cornicles about two-thirds as long as VI spur and slightly shorter than III,

the latter with about 15 sensoria (fig. 133) ................ cucurbitae (Thomas)

20. Ground or basal color of abdomen green ............................................................ 21

— Ground or basal color of abdomen red, brown, or black .................................... 24

21. Cornicles green, sometimes dusky at apex ........................ solanifolii (Ashmead)

— Cornicles black ........................................................................................................ 22

22. Ill with a small number (9-15) of secondary sensoria on basal one-half

(fig. 135) ; longer than VI spur ........................................ granarium (Kirbyj

— Ill with some 30 or more sensoria scattered along its entire length (figs. 151,

159); subequal to or shorter than VI spur ......... . ........................................ 23

23. Cornicles and III subequal. Tibiae with apices only dusky ........ rosae (Linn.)

— Cornicles longer than III. Tibiae dusky throughout.

rudbeckiae (Fitch) n.var. madia

24. Cauda pale ................................................................................................................ 25

— Cauda dusky ............................................................................................................ 27

25. Ill and VI spur subequal ................... : ................................................ rosae (Linn.)

— Ill shorter than VI spur .............................................. . ....................................... 26

26. Cauda about one-half as long as cornicles, the latter shorter than IV.

chrysanthemi (Oestlund)

— Cauda slightly more than one-half as long as cornicles, the latter equal to or

longer than IV ...................................................................... rudbeckiae (Fitch)

27. Ill and VI spur subequal ...................................................................................... 28

— Ill longer than VI spur .................................................................. taraxici (Kalt.)

28. Body yellowish-brown in color; legs same except tarsi and tips of tibiae and

femora which are dusky to black ...................................... valerianae (Clarke)

— Body dark reddish-brown to black in color; legs dusky throughout.

ambrosiae (Thomas)

Apterous viviparous
1. Cornicles clavate on one side, somewhat as in Elwpalosiplium.

tulipae (Monell)
— Cornicles not so, being cylindrical or subcylindrical .......................................... 2

10 Only the species of which there are specimens available to the author, or of
which there are adequate descriptions, are included in this key. The species rep-
resented in the author's collection are marked with an asterisk (*). The author
recognizes the great difficulty in separating the apterae of various species, par-
ticularly in this genus, and offers this key merely as a slight aid toward the recog-
nition of the better known species.

A SYNOPSIS OF THE APHIDIDAE 59

2. Ill without or at most with only a few secondary sensoria (0-12) 11

— Ill with several (over 12) secondary sensoria scattered along the greater part

of its length 3

3. Cornicles short and tapering, being somewhat bottle-shaped and not distinctly

longer than the cauda sanborni Gillette*

— Cornicles normal, being cylindrical and considerably longer than the cauda 4

4. Ill and IV with secondary sensoria heucherae (Oestlund)

— IV without secondary sensoria 5

5. General body color dark, being red, wine, brown or black 6

— • General color lighter, usually being a shade of green 8

6. Cauda black. Legs black, except the bases of the femora taraxici (Kalt.)

— Cauda pale. Legs with at least the bases of the femora and tibiae not black 7

7. Legs green, except tarsi and apices of femora and tibiae. Cauda not more than

half the length of the cornicles. Not more than ten to twelve sensoria on
the basal one-third of III rosae (Linn.)*

— Legs black, except bases of femora and tibiae, which are light brown. Cauda

more than half the length of the cornicles. A considerable number of
sensoria scattered over more than the basal one-half of III.

rudbeckiae (Fitch)*

8. Cornicles subequal to or shorter than III. Body covered with a whitish pul-

verulence 9

— Cornicles distinctly longer than III. Body without whitish pulverulence .... 10

9. Cornicles, except tip, and cauda green; the former subequal in length to III

and about twice as long as cauda albifrons Essig*

— Cornicles black, cauda yellow or light brown; the former considerably shorter

than III and not twice as long as cauda ludovicianae (Oestlund)*

10. Cauda quite broad and blunt at end. Cornicles with not more than apical one-

sixth reticulated rosae (Linn.) *

— Cauda slender and pointed. Cornicles with apical one-fourth reticulated.

rudbeckiae (Fitch) n.var. madia*

11. Body covered with capitate or fan-shaped setae 12

— Body without specialized setae 14

12. Setae with fan-shaped tips and thickly covering the body. Cornicles slender

and imbricated for their entire length artemisicola (Williams)*

— Setae capitate and only sparsely covering body 13

13. Cornicles fairly stout, with tips reticulated, and about twice as long as cauda.

artemisiae (Fonsc.)

— Cornicles slender, with no reticulations, and considerably more than twice the

length of the cauda pteridis Wilson

14. Cornicles with tips at least reticulated 16

— Cornicles with no reticulations 15

15. Cornicles very long and slender. Antennae considerably longer than body.

pisi (Kalt.)*

— Cornicles shorter and heavier. Antennae at most but slightly longer than

body dirhodum (Walker)*

16. Cornicles for the most part dusky or black 17

— Cornicles mostly pale or green 19

17. Cornicles and III subequal. Body not pulverulent 18

— Cornicles considerably shorter than III. Body more or less pulverulent.

ludovicianae (Oestlund)*

60 MISCELLANEOUS STUDIES

18. Ill with but two or four sensoria near base; longer than VI spur.

gr anarium ( Kirby ) *

— Ill with six or so sensoria on basal one-half ; shorter than or equal to VI

spur rosae (Linn.)*

19. Cornicles longer than III 20

— Cornicles at most subequal to III 21

20. Antennae pale, except VI and the apices of III to V. Cornicles slightly

swollen near distal end Stanley! Wilson*

— Antennae dusky, except III, basal part of IV, and perhaps the extreme base

of V. Cornicles long, slender, and out-curved californicum (Clarke)*

21. Cauda broad, and blunt, with the sides almost parallel and about half as long

as cornicles lactucae (Kalt.) *

— Cauda slender-pointed, and more than half as long as cauda 22

22. VI spur and III subequal solanifolii (Ashmead)*

— VI spur considerably longer than III cucurbitae (Thomas)*

59. Macrosiphum albifrons Essig

Figures 143, 144

Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909. Macrosiphum sp. (list).
Essig, Pom. Jour. Ent., vol. 3, p. 543, 1911 (orig. desc.).

Records. — Lupinus sp., Santa Paula (Essig) ; Stanford University (Davidson) ;
Jasper Ridge, Coast Range Mountains, Santa Clara County, April, 1912 (V. G.
Stevens) ; Berkeley, April, 1915 (Geo. Shinji) ; Mount Hood, Oregon, August,
1916 (E. A. McGregor).

This large, flocculent aphid is found occasionally infesting various
lupines throughout the Pacific Coast, from southern California north,
well into Oregon. The author has specimens from Berkeley and
Oregon, although he has never collected it himself.

60. Macrosiphum ambrosiae (Thomas) ?

Thomas, 111. Lab. Nat. Hist., Bull. 2, p. 4, 1878. Siphonoplwra (orig.

desc.).
Sanborn, Kans. Univ. Sci., Bull. 3, p. 74, 1904 (desc.).

Beoords. — Helianthus annuus; Orange (T. D. A. Cockerell) ; San Diego, April,
1916.

In 1915 the author received a few specimens of this species from
T. D. A. Cockerell from Orange, and in 1916 he collected it once on
sunflower in Exposition Park, San Diego. At first it was thought to
be M. sonchi (Linn.), and was so reported by Cockerell. Since then
it was identified by J. J. Davis as probably M. ambrosiae (Thomas).

A SYNOPSIS OF THE APHIDIDAE 61

61. Macrosiphum artemisiae (Fonsc.)

Figures 142, 145

Boyer de Fonscolmbe, Ann. Ent. Soe. France, vol. 10, p. 162, 1841. Aphis

(orig. desc.).
Essig, Pom. Jour. Ent., vol. 3, p. 546, 1911. Macrosiphum frigidae (Oest.)

(desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 133, 1914. Macrosiphum frigidae

(Oest.) (list).
Wilson, Trans. Amer, Ent. Soc., vol. 41, p. 97, 1915 (desc.).

Records. — Artemisia calif ornica ; Santa Paula (Essig) ; Walnut Creek, Contra
Costa County (Davidson).

Occasionally this species is found infesting the tender shoots of
the common California sage brush. It is characterized by the presence
of capitate hairs scattered sparsely over the body, particularly of the
apterous female. The synonomy above is after Wilson, who lists
M. frigidae (Oestlund) as a synonym of artemisiae (Fonsc.).

62. Macrosiphum artemisicola (Williams)

Figures 146, 147

Williams, Univ. Neb. Studies, vol. 10, p. 73, 1910. Siphonophora (orig.

desc.).
Wilson, Trans. Am. Ent. Soe., vol. 41, p. 96, 1915 (desc.).

Records. — Artemisia tridentata, A. vulgaris; Oregon (California) (Wilson).

Although there is no published record of the presence of this
species in California it is included here on Wilson's authority. He
stated to the author that he had found it in California, although he
failed to give any date or locality record. This is characterized by
the fan-shaped setae which thickly cover the body of the apterae, and
which are present on the ventral side of the abdomen of the alates.
The author has specimens taken by R. W. Haegele in the summer of
1915 on Artemisia sp. near Canton, Montana.

63. Macrosiphum baccharadis (Clarke)

Clarke, Can. Ent., vol. 35, p. 254, 1903. Nectarophora (orig. desc.).
Record. — Baccharis sp., Berkeley (Clarke).

This species is one of those described by Clarke, but since then
unknown. It is possible that it is M. rudbeckiae (Fitch), which is so
common on Baccharis throughout California.

62 . MISCELLANEOUS STUDIES

64. Macrosiphum calif ornicum (Clarke)

Figures 131, 132

Clarke, Can. Ent., vol. 35, p. 254, 1903. Nectarophora (orig. desc. apterae).
Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910 (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911 (list).
Essig, Pom. Jour. Ent., vol. 3, p. 548, 1911. M. laevigatae, n.sp. (orig.
desc.).

Eecords. — Salix sp. ; Newcastle, Placer County (Clarke); Stanford University
and Penryn, Placer County (Davidson) ; Stanford University, November, 1914
(Morrison), May, 1915; Berkeley, April, 1915 (Shinji) ; August, 1915, Salix
laevigata; Santa Paula (Essig) ; Riverside, May, 1917.

Clarke described the apterous females of a species of Nectarophora
(Macrosiphum) from specimens taken on willow in Placer County.
Because of the extremely long cornicles it is possible to identify this
with specimens taken since throughout the San Francisco Bay region
on various species of willows. Essig 's M. laevigatae from Santa
Paula is the same species, having been compared by the author with
specimens from Stanford University and Berkeley. Morrison has
taken the males and oviparous females of this species in the vicinity
of Stanford University in November, 1914. The author has reared
specimens of Aphidins polygonaphis Fitch, and Praon simulans Prov.
from this species taken in Berkeley.

65. Macrosiphum chrysanthemi (Oest.)

Oestlund, 14th Rep. Geol. Surv. Minn., vol. 22, 1886. Siphonophora (orig.

desc.).

Davidson, Jour. Econ. Ent., vol. 5, p. 411, 1912 (list).

.Record.— Undetermined species of Compositae; Courtland (Davidson).

This is a doubtful species taken by Davidson at one time from an
undetermined composite near Courtland. The author is entirely
unacquainted with the species.

66. Macrosiphum cucurbitae (Thos.)

Figures 133, 134

Thomas, 8th Ann. Rep. Illinois St. Ent., p. 66, 1879. Siphonophora (orig.
desc.).

Becord. — Cucur'bita sp., Hayward, Alameda County, July, 1915 (Roy E. Camp-
bell) ; Los Angeles, May, 1917.

A SYNOPSIS OF THE APHIDIDAE 63

In July, 1915, Roy E. Campbell of the Bureau of Entomology,
sent the author specimens of a Macrosiphum sp. from squash in Hay-
ward. In 1917 the author found the same species abundantly on
squash in Los Angeles. These the author identified as being specimens
of M. cucurbitae (Thomas). Later J. J. Davis verified the deter-
mination. This is a new record for California. As the available
descriptions of this species are quite inadequate, the author gives
herewith a few descriptive notes taken from these specimens..

Alate viviparous female. — Antennae longer than the body, placed
on distinct frontal tubercles, dusky except I, II, and extreme base
of III. The spur of VI is the longest segment, followed by III, which
is about four-fifths as long. IV and V are subequal, and almost as
long as III. The usual primary and accessory sensoria are present
on V and VI. Secondary sensoria are present on III (fig. 133), being
small, circular, numbering about 14 to 15, and arranged in a fairly
even row along the whole length of the segment. Beak pale with
dusky tip, reaching to the second coxae. Thorax and abdomen green,
the thoracic lobes not conspicuously darkened. Cornicles (fig. 134)
green with apical one-third dusky, equal to or slightly longer than
III, imbricated with tip reticulated. Cauda large, pale, vasiform,
slightly more than half the length of the cornicles, reaching to their
apices. Wings and legs normal.

Measurements : Body length, 2.3 mm. ; antennae total, 3.25 to 3.35
mm. ; III, 0.685 to 0.714 mm. ; IV, 0.629 to 0.646 mm. ; V, 0.603 to
0.612 mm. ; VI, base, 0.136 to 0.153 mm. ; VI, spur, 0.935 to 0.696 mm. ;
cornicles, 0.714 to 0.731 mm. ; cauda, 0.408 mm.

67. Macrosiphum dirhodum (Walker)

Figures 156, 157

Walker, Ann. Nat. Hist., (2), vol. 3, p. 43, 1848. Aphis (orig. desc.).
Theobald, Jour. Econ. Biol., vol. 8, p. 128, 1913 (desc.).
Patch, Maine Agr. Exp. Sta., Bull. 233, p. 268, 1914 (note).
Gillette, Jour. Econ. Ent., vol. 8, p. 103, 1915 (note).

Record. — Eose, Santa Ysabel (3000 feet altitude), San Diego County, May,
1916; Eiverside, April, 1917.

The author found this species sparingly on rose near Santa Ysabel,
San Diego County, in May, 1916, and again in April, 1917, in River-
side. According to Gillette, this species passes the winter on rose,
and the summer on various grains and grasses, as M. rosa-e (Linn.)

64 MISCELLANEOUS STUDIES

may do. These are the only records of it in California. The author
has compared it with specimens taken by R. W. Doane in 1915 on
grain in Utah.

68. Macrosiphum granarium (Kirby)
Figures 135, 148

Kirby, Linn. Soc. Trans., vol. 4, p. 238, Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 5, p. 411, 1912 (list).
Theobald, Jour. Econ. Biol., vol. 8, p. 58, 1913 (desc.).
Davidson, Mon. Bull., Cal. Comm. Hort., vol. 6, p. 65, 1917 (note).

Records. — Graminaceae (various species) ; San Jose (Davidson) ; Stanford
University, January to May, 1915 ; Berkeley, March, 1915 : Typha latifolia
(Davidson).

This is a more or less common species of Macrosiphum on various
grains and grasses in the San Francisco Bay region during the winter
and spring. In late spring and early summer, as the grasses begin
to dry out, it leaves them for the cat-tail rush or California tule
(Davidson, 1917). In the late fall or early winter it returns to the
grains and grasses, where it passes the winter in the viviparous forms.

69. Macrosiphum heucherae (Thomas)

Thomas, 8th Ann. Kep. Illinois St. Ent., p. 66, 1879. Siphonophora (orig.

desc. ) .
Davidson, Jour. Econ. Ent., vol. 8, p. 427, 1915 (desc.).

Record. — Heuchera ihartwegi, Kedwood Canyon, Contra Costa County (David-
son).

In the latter part of May, 1914, Davidson found all the forms,
including the apterous and alate viviparous females, the apterous
oviparous females, the alate males, and eggs on the flower stalks of
alum root in Contra Costa County. Since his description no record
has been made concerning the species. The author is unacquainted
with it, having never seen specimens.

70. Macrosiphum jasmini (Clarke)

Clarke, Can. Ent., vol. 35, p. 252, 1903. Ncctarophora (orig. desc.).
Record. — Jessamine, Berkeley (Clarke).

Since Clarke's description of the apterous viviparous females of
this species it has never been found. Its identity is, therefore, un-
known to the author.

A SYNOPSIS OF THE APHIDIDAE 65

71. Macrosiphum lactucae (Kalt.)

Kaltenbach, Monog. d. Pflanzenlause, p. 199, 1857. Nectarophora (orig.

desc.).

Sanderson, Can. Ent., vol. 33, p. 69, 1901. Nectarophora (desc.).
Essig, Univ. Calif. Publ., Entom., vol. 1, p. 328, 1917 (list).

Record. — Cicorium intybus, Kutherford^ Napa County, 1916 (Essig).

This species has been taken only by Essig on chicory in Napa
County during June, 1916. As its determination is doubtful the
author gives herewith a brief description of the alate female.

Body pale to green, with the following parts more or less dusky :
head, antennae, prothorax, thoracic lobes, apex of beak, tarsi, apical
one-fifth to one-fourth tibiae, apical one-half femora, cornicles, anal
plate, marginal spots on the abdominal segments, submarginal spots
of the second and third abdominal segments, dorsal bands on the
fourth and fifth, and the dorsum of the remaining abdominal seg-
ments. Eyes red.

The antennal tubercles are prominent and project rectangularly
inward. A prominent frontal tubercle is present on the apex. The
antennae are about half as long again as the body. The usual primary
and accessory sensoria are present. On III there are from thirty-five
to forty-five circular secondary sensoria; on IV from five to fifteen
secondary sensoria. These two segments appear tuberculate. The
beak reaches beyond the second coxae. The cornicles are longer than
the cauda, and subequal in length to the fourth antennal segment.
They are subcylindrical and fairly stout. The cauda is long and ensi-
form, reaching to the tip of the cornicles. The wings and venation
are normal.

Measurements (of three specimens) : Body length, 1.836 to 1.955
mm. ; width of thorax, 0.765 to 0.833 mm. ; antennae, total, 2.805 to
2.992 mm. ; III, 0.680 to 0.697 mm. ; IV, 0.441 to 0.527 mm. ; V, 0.391
to 0.425 mm. ; VI, base 0.085 to 0.119 mm. ; VI, spur 0.952 to 1.105
mm. ; cornicles, 0.441 to 0.493 mm. ; cauda, 0.238 to 0.272 mm. ; hind
tarsi, 0.136 to 0.153 mm. ; wing, length, 3.145 to 3.315 mm. ; width,
0.952 to 1.139 mm. ; expansion, 7.36 to 7.87 mm.

72. Macrosiphum ludovicianae (Oestund)

Figures 136, 148

Oestlund, Minn. Geol. Nat. Hist. Surv., vol. 14, p. 23, 1886. Siphonophora

(orig. desc.).

Davidson, Jour. Econ. Ent., vol. 7, p. 136, 1914 (list).
Wilson, Trans. Am. Ent. Soc., vol. 12, p. 98, 1015 (desc.).

66 MISCELLANEOUS STUDIES

Records. — Artemisia heterophylla ; Walnut Creek, Contra Costa County (David-
son), Berkeley, 1915 (Shinji); Artemisia dracunculoides, Convolvulus sp., Stachys
bullata, Berkeley, 1915 (Shinji).

This species is quite common in the San Francisco Bay region on
various species of sagebrush. George Shinji has taken it also on hedge-
nettle and bindweed in Berkeley. It is distinguished from other sage-
infesting species of Macrosiphum by the fact that the body of the
apterous females is covered with pointed setae as opposed to the fan-
shaped setae of M. artemisicola (Williams), and the capitate setae
of M. artemisiae (Fonsc.).

73. Macrosiphum orthoearpus Davidson

Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909 (orig. dese.).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910 (list).

Record. — Orthoearpus purpurascens ; Stanford University (Davidson).

Since Davidson found the specimens on owl-clover from which he
described this species, it has not again been taken.

74. Macrosiphum pisi (Kalt.)

Figures 130, 150

Kaltenbach, Monog. d. Pflanzenlause, p. 23, 1843. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910 (list).
Essig, Pom. Jour. Ent., vol. 2, p. 336, 1910. Nectarophora (desc.).
Branigan, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 285, 1915. M. destruc-
tor (Johnson) (list).
Davis, U. S. Dept. Agr., Bull. 276, p. 11, 1915 (list).

Records. — Pisum sativum; Claremont, Santa Ana, and Ventura (Essig) ; Ala-
meda County (Brannigan) ; El Cajon, San Diego County, May, 1916: Lathyrus
odoratus; Stanford University (Davidson, Morrison); San Diego, October, 1916:
Viola sp. ; Claremont, Santa Ana, Ventura (Essig); Medicago sp. ; Holtville,
Imperial County (V. L. Wildermuth) : Psorales macrostachya; Santa Paula
(Essig).

The pea aphis is quite common throughout the state, especially on
garden and sweet peas. It has been taken a few times on other plants,
such as alfalfa, violets, and leather-root, but it is uncommon. This
species is readily distinguished by its bright, shining green color, large
size, and long, slender, imbricated, but non-reticulated cornicles.

A SYNOPSIS OF THE APHIDIDAE 67

75. Macrosiphum pteridis Wilson

Figures 317, 318
Wilson, Trans. Am. Ent. Soc., vol. 41, p. 101, 1915 (orig. desc.).

Records. — Pteris aquilina; Walnut Creek, Contra Costa County, 1915 (David-
son).

This species has been found by Davidson on the fronds of common
brake in the San Francisco Bay region. Wilson reported it as present
throughout southern and western Oregon. There are a few specimens
of the alate females in the author 's collection, received from Davidson.

76. Macrosiphum rosae (Linn.)

Figures 106, 151, 152

Linnaeus, Syst. Nat. vol. 4, p. 73, 1735. Aphis (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 254, 1903. Nectarophora (list).

Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909 (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910 (list).

Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911 (list).

Essig, Pom. Jour. Ent., vol. 5, p. 550, 1911 (desc.).

Games, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 398, 1912 (list).

Records. — Eose; through California from Humboldt County south to San Diego
County (Clarke, Davidson, Morrison, Essig, Ferris, Shinji, the author).

This is the common pink and green aphid of roses, known the
world over. The apterae are found most abundantly in the late win-
ter and early spring on the buds and stems of rose. As the alates are
matured they fly away, supposedly either to other rose bushes or to
various grains and grasses. This past spring (1917) it has been very
abundant in the vicinity of Eiverside, but the previous spring (1916)
in San Diego it was rare. There the most abundant rose aphis was
Myzaphis rosarum (Walker).

77. Macrosiphum rudbeckiae (Fitch)

Fitch, Cat. Homop. N. Y., p. 66, 1851. Aphis (orig. desc.).
Essig, Pom. Jour. Ent., vol. 3, p. 400, 1911. Aphis (dese.)."
Davidson, Jour. Econ. Ent., vol. 7, p. 137, 1914 (list).

Records. — Ambrosia psilostachya; Santa Paula (Essig) ; Baccharis viminalis;
Santa Paula (Essig), Riverside, September, 1916; Dipsacus fullonum; San Jose
(Davidson): Helianthus annuus; Eiverside, September, 1916; Salix sp. ; Chrysan-
themum; Arlington, Eiverside County, September, 1916; undetermined species of
Compositae; Bedwood Canyon, Contra Costa County, July, 1914 (E. W. Haegele).

11 In the drawings accompanying this description by Essig the following mis-
takes are noticeable: the third discoidal vein of the forewings is twice-branched
instead of once-branched, and the third antenal segment of the apterous female
bears several secondary sensoria instead of none, as figured.

68 MISCELLANEOUS STUDIES

This reddish-colored Macrosiphum is distributed abundantly
through the San Francisco Bay region and southern California on
various Compositae. In one case the author found it doing consider-
able damage to chrysanthemums by stunting and distorting the buds.
Once he found it infesting the tender leaves and stalks of willow.
The author reared specimens of Diarctns rapae Curt, from an infesta-
tion of this species taken on willow.

77a. Macrosiphum rudbeckiae (Fitch) var. madia n.var.

Figures 153, 154

In September, 1915, the author found a species of Macrosiphum
infesting the heads of tarweed (Madia sativa) on the campus of the
University of California, Berkeley. Specimens of Praon simulans
Prao. were reared from this collection. Mounted specimens are almost
identical with M. rudbeckiae (Fitch), but in life they differ in the
coloration. Because of this it has been thought best to describe it
herewith as a color variety of M. rudbeckiae, naming the variety.
madia, after its host plant.

Host: Madia sativa. Date: September 12, 1915.

Locality: Berkeley, California. Collection number: AFS 70-15.

Alate viviparous female. — Prevailing color: dark-green, slightly
pruinose. Head brownish (fuscous), about as long as broad, with
distinct antennal tubercles. Antennae black, except I and II and the
base of III, which are concolorous with the head. The spur is slightly
longer than III ; IV is next in length, followed by V, VI, and I, which
are subequal, and II, which is the shortest segment. The spur is about
six times as long as the base of VI. The usual primary sensoria are
present on V and VI, and the usual accessory sensoria on VI. IV is
without sensoria, III has 25-35 irregularly arranged, various-sized
secondary sensoria placed along the whole length of the segment
(fig. 154). The thorax is fuscous; the prothorax with rather distinct
lateral tubercles. The beak is slightly dusky with the apical one-third
black, reaching to the second coxae. The abdomen is greenish with a
slight pulverulence, making it appear pruinose. The cornicles are
long, slightly tapering, black except the basal one-third, which is
concolorous with the abdomen, apical one-fifth reticulate (fig. 153).
The cauda is long and pointed, pale (slightly reddish?), about one-
half as long as the cornicles. The legs are black except the basal half
of the femora and the coxae, which are greenish. The wings and
venation are normal.

A SYNOPSIS OF THE APHIDIDAE 69

Measurements: Body length (exclusive of cauda), 2.11 mm. ; width
of thorax, 0.91 mm. Antennae : total, 2.07 mm. ; I, 0.12 mm. ; II,
0.09 mm. ; III, 0.76 mm. ; IV, 0.55 mm. ; V, 0.47 mm. ; VI, 0.12 mm. ;
spur, 0.78 mm. ; cornicles, 0.91 mm. ; cauda, 0.45 mm. ; beak, 0.89 mm. ;
hind tarsus, 0.14 mm. Wing: length, 3.6 mm.; width, 1.25 mm.;
expansion, 8.11 mm.

78. Macrosiphum sanborni Gillette

Figures 141, 155

Sanborn, Kans. Univ., Sci. Bull. 3, p. 73, 1904. Macrosiphum chrysanthemi

(desc. ala. vivi.).
Gillette, Can. Ent., vol. 11, p. 65, 1908 (orig. desc. apt. vivi.).

Records. — Chrysanthemum; Stanford University, May, 1915; Riverside, March,
1917.

Twice has the author found this species: once a small infestation
in the greenhouse of Stanford University, and once abundantly out
of doors in Riverside. It is an interesting species in that it does not
fit well into any known genus. Except for the cornicles it fits Macro-
siphum and has been so considered. The cornicles are, however, short,
being scarcely longer than the cauda, and are somewhat bottle-shaped,
being considerably smaller at the apex than at the base.

79. Macrosiphum solanifolii (Ashmead)

Figures 137-140, 159-160

Ashmead, Can. Ent., vol. 12, p. 91, 1881. Siphonophora (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 252, 1903. Nectarophora citrifolii (Ashmead)
(list).

Davidson, Jour. Econ. Ent., vol. 31, p. 380, 1910. Macrosiphum citrifolii
(Ashmead) (list).

Essig, Pom. Jour. Ent., vol. 3, p. 592, 1911. Macrosiphum citrifolii (Ash-
mead) (desc.).

Davidson, Jour. Econ. Ent., vol. 5, p. 411, 1912 (list).

Patch, Maine Agr. Exp. Sta., Bull. 242, 1915 (desc.).

Records. — Citrus sp. ; Azusa, Los Angeles County (Clarke); Lindsey, Tulare
County (Clarke) ; Santa Paula (Essig) ; Disporum Tnookeri; Berkeley, May, 1915
(Shinji) : Solatium nigrum; Stanford University, October, 1916 (Ferris) : Fuchsia
sp. ; Berkeley, July, 1915: Sonohus asper and S. oleraoeus; Stanford University,
February, 1915: apple; Stanford University, May, 1915; El Cajon, San Diego
County, July, 1916: Atriplex sp. ; Berkeley, September, 1915: Oxalis corniculata,
Eiverside, February, 1917: Deinandra fasciculata, Eiverside, February, 1917:
Erodium moschatum; Pasadena, April, 1917 (E. E. Campbell); Eiverside, April,
1917.

70 MISCELLANEOUS STUDIES

This "pink and green aphid of potato" is distributed throughout
California on a large variety of plants. It is recognizable by the long
reticulated cornicles and black antennae. When the author first exam-
ined specimens of Macrosiphum citrifolii (Ashmead) in Essig's collec-
tion he was struck with its resemblance to this species. In fact, after
considerable study he could not find any constant differences. This was
in 1915 in Berkeley. This past spring (1917) he had the opportunity
in Riverside of making some transfer tests with specimens from oxalis.
Migrants were placed under muslin bags on sucker growth of orange.
It was observed that these settled there readily and produced young,
demonstrating that the citrus species is the same as the other. On
the strength of this Macrosiphum citrifolii (Ashmead) is listed as a
synonym of this species.

80. Macrosiphum sonchella (Monell) ?

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 21, 1879. Siphonophora (orig.

desc.).

Clarke, Can. Ent., vol. 35, p. 252, 1903. Nectarophora (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 380, 1910 (list).

Records. — Sonchus sp. ; Berkeley, Newcastle, and Palo Alto (Clarke); Stan-
ford University (Davidson).

According to Morrison the species listed as this by Davidson is
not Macrosiphum sonchella (Monell), although he cannot say what
it is. Consequently Clarke probably referred to the same species as
did Davidson. As the author has never seen specimens he can make
no statement as to its identity, so lists it as it has been heretofore.

81. Macrosiphum Stanley! Wilson

Figures 128, 158

Wilson, Proc. Ent. Soc. Brit. Columbia, January, 1915 (orig. desc.).
Eecord. — Sambiuxis callicarpa californica; Berkeley, June, 1915.

From the early part of June, 1915, until the middle of August,
this species was very abundant on an elderberry tree in the Botanical
Gardens of the University of California. By the latter part of August
all specimens had disappeared. Since then the author has never seen
the species. J. J. Davis kindly identified these specimens.

A SYNOPSIS OF THE APHIDIDAE 71

82. Macrosiphum taraxici (Kalt.)

Kaltenbach, Monog. d. Pflanzenlause, p. 30, 1743. Aphis (orig. desc.).
Theobald, Jour. Econ. Biol., vol. 7, p. 77, 1913 (desc.).

Record. — Taraxacum officinale; California (Wilson).

H. F. Wilson stated to the author that he had taken this species
on dandelion (Taraxacum officinale} in California, although he gave
no date or locality record.

83. Macrosiphum tulipae (Monell)

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 19, 1879. Siphonophora (orig.

desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910 (list).

Records. — Tulipa sp. ; Stanford University (Davidson); Liriodendron sp. ;
Berkeley, 1915 (Essig, Shinji).

This species is not known to the author. It has been found on
tulips and on the tulip trees in the San Francisco Bay region by
Davidson, Essig, and Shinji.

84. Macrosiphum valerianae (Clarke)

Clarke, Can. Ent., vol. 35, p. 253, 1903. Nectarophora (orig. desc.).
Record. — Valeriana offlcinialis; Berkeley (Clarke).

In 1903 Clarke described this species from specimens taken on
heliotrope in Berkeley. Since then it has not again been found.

24. Genus Myzus Passerini

Passerini, Gli Afidi, 1860. Type Aphis ribes Linnaeus.
This genus is very closely related to Rhopalosiphum Koch, the
principal difference being in the shape of the cornicles. However,
some species fall easily into one or the other genus, depending entirely
upon what form one has. In this respect Rhopalosiphum persicae
(Sulz.) is particularly noticeable, the spring migrants having the
clavate cornicles of Rhopalosiphum, the fall migrants having the
cylindrical cornicles of Myzus. The author has followed Van der
Goot in taking out of this genus M. rosarum (Walker) and placing it
in the genus Myzaphis v.d.G. The antennal tubercles are lacking,
thus placing the species in the Aphidini instead of the Macrosiphini.
There are at present ten species of Myzus known to occur in Califor-
nia. Following is a key to them:

72 MISCELLANEOUS STUDIES

KEY TO THE CALIFORNIA SPECIES
Alate viviparous females

1. Secondary sensoria present on III only 5

- Secondary sensoria present on other segments as well as on III 2

2. Secondary sensoria on III, IV, and V 3

- Secondary sensoria on III and IV, none on V fragaefolii Cockerell

3. Cornicles dusky for entire length cynosbati (Oestlund)

— Cornicles mostly pale 4

4. Thoracic lobes distinctly darker than general body color, being black or dark

brown aquilegia Essig

— Thoracic lobes at most only slightly darker than body, being a pale brown.

braggli Gillette

5. Body black throughout cerasi (Fabricius)

— Body not black throughout 6

6. Cornicles pale except at extreme tip 7

— Cornicles with more than tip dusky 8

7. VI spur longer than III, the latter with but 9 to 12 sensoria.

varians Davidson

— VI spur at most equal to III, the latter with 18 to 26 sensoria.

lycopersici (Clarke)

8. Cornicles longer than either IV or V 10

— Cornicles not longer than either IV or V 9

9. VI spur longer than II circumflexum (Buckton;

— VI spur shorter than III rlbifolii Davidson

10. Ill with 15 to 25 sensoria (fig. 178) rhamni (Fonsc.)

- Ill with but 9 to 12 sensoria varians Davidson

Apterous viviparous female

1. Body covered with capitate hairs 2

— Body not covered with capitate hairs except on head and antennae 5

2. Secondary sensoria on III 3

— No secondary sensoria on III 4

3. Cornicles dusky ribifolii Davidson

— Cornicles pale except tip aquilegiae Essig

4. Cornicles almost twice as long as III. Body fairly large sized.

braggii Gillette

— Cornicles but slightly longer than III. Body small sized.

fragaefolii Cockerell

5. Secondary sensoria on III 6

— No secondary sensoria on III 7

6. VI spur longer than III. Several sensoria scattered along the whole length

of III cynosbati (Oestlund )

— VI spur at most equal to III. Only a few (1-3) sensoria at base of III.

lycopersici (Clarke)

7. Ill longer than cornicles. Dorsum of abdomen with dusky markings, shaped

somewhat as a horseshoe circumflexum (Buckton)

— Ill at most equal to cornicles. Abdomen not marked as above 8

8. Body black throughout cerasi (Fabricius)

— Body not black throughout 9

9. VI spur almost twice as long as III varians Davidson

— VI spur but slightly longer than III rhamni (Fonsc.)

A SYNOPSIS OF THE APH1DIDAE 73

85. Myzus aquilegiae Essig

Shinji, Can. Ent., vol. 40, p. 49, 1917. Myzus sp. (list).

Essig, Univ. Calif. Publ Entom., vol. 1, p. 314, 1917 (orig. desc.).

Eecords. — Aquilegia truncata; Berkeley, 1916 (Essig) : A. vulgare, Inverness,
Marin County (Shinji).

This species was recently described by Essig from specimens found
on columbine on the campus of the University of California, Berkeley.
The author has had access to cotype specimens, although he has never
collected it himself.

86. Myzus braggii Gillette

Figure 176

Gillette, Can. Ent., vol. 11, p. 17, 1908 (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 5, p. 409, 1912. Phorodon carduinum
(Walker) (list).

Eecords. — Cynara scolymus; Courtland, Oakland, and San Jose (Davidson) ;
Riverside, January and February, 1917.

The author found this species during the early spring of 1917
infesting the leaves of artichoke in Riverside. The determination of
specimens was verified by C. P. Gillette. Davidson reported Phorodon
carduinum (Walker) from artichoke in the San Francisco Bay region.
His specimens were determined by J. Monell, but P. Van der Goot
was doubtful as to its identity. Davidson himself has decided that
the species is Myzus braggii Gillette. There is no doubt but that the
species on artichoke in California is M. braggii Gillette, but whether
or not this is the same as P. carduinum (Walker) is uncertain.

87. Myzus cerasi (Fabricius)
Figures 112, 121, 122, 179, 307

Fabricius, Syst. Nat., p. 734. Aphis (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 252, 1903 (list).

Gillette, Jour. Econ. Ent., vol. 1, p. 362, 1908 (desc.).

Newman, Mon. Bull. Calif. Comm. Hort, vol. 4, p. 446, 1915 (list).

Shinji, Can. Ent., vol. 49, p. 49, 1917 (list).

Eecords. — Prunus cerasi; Susanville, Lassen County (Newman) ; Berkeley,
1914, 1915, and 1916 (Essig, Shinji) ; Eiverside, 1914 (Sharp) ; Fresno, June,
1915: Prunus domestica; Berkeley (Clarke).

74 MISCELLANEOUS STUDIES

The black cherry aphis is found occasionally throughout Califor-
nia, but seldom in large enough numbers to be injurious. It infests
the terminal leaves of cherry, and sometimes other species of Prunus,
causing them to curl to a certain extent. Eggs are laid in the late fall
and early winter in the crevices of the bark and near the bases of the
buds. These hatch the following spring about the time the buds are
opening. The first few generations consist entirely of apterous
females. In the early summer the alate females appear, and con-
tinue to do so in each succeding generation until fall. In fact, after
the first of July, or thereabouts, the majority of the lice produced
are alate until the sexes appear in the fall. The first alate females
taken by the author were on June 7, 1915. However, on April 25,
1916, Essig found a few alate females in Berkeley. In August, 1914,
the apterae were also found in Berkeley.

Van der Goot makes this species out of the genus Myzus, using it
as the type of his genus Myzmdes. The author is inclined to follow
him inasmuch as this is quite different from other members of this
genus, approaching Aphis in its robust form and separated from
that only by the length of the cornicles and presence of antenna!
tubercles. However, it has so long been considered as a species of
Myzus that it is best to leave it so. It is not a good policy usually to
form a new genus for one species, especially when it has for so long
been considered as a member of another genus.

88. Myzus circumflexus (Buckton)

Figure 175

Buckton, Monog. Brit. Aphides, vol. 1, p. 130, 1875. Siphonopliora (orig.

desc.).

Gillette, Can. Ent., vol. 40, p. 19, 1908. M. vincae, n.sp. (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 380, 1910. M. vincae Gill. (list).
Shinji, Can. Ent., vol. 49, p. 49, 1917 (list).

Record. — Vinca major; Stanford University (Davidson, Morrison), Berkeley,
1915 (Shinji), Los Angeles, March, 1917; Aesculus californicus, Alopecurus
pratensis, Asparagus, spp., Ceanothus sp., Cerastium viscosum, Cheiranthus chieri,
Cyrtonium falcatum, Digitalis purpurea, Fuchsia sp., Gladiolus sp., Plantago sp.,
Senecio mikanioides, Sisymbrium sp., Solanum spp., Stachys bullata, Tropaeolum
sp., Symphoricarpus racemosus; Berkeley, 1915, 1916 (Essig, Shinji) : Viola tri-
color; Stanford University, March, 1915; Berkeley, 1915 (Essig, Shinji):
Rioliardia africana; Pomona, 1909 (Essig); Stanford University, March, 1915;
Berkeley, March, 1915 (Essig) ; San Diego, May, 1916; Los Angeles, March, 1917.

This very common aphid is found in the spring on a large variety
of host plants throughout California. At times it may become so

X

A SYNOPSIS OF THE APHIDIDAE 75

abundant as to cause some considerable damage to its host. On March
4, 1917, the author observed it on periwinkle in Los Angeles in such
numbers as to stunt the flowers and to cause all the plants to appear
black and sticky. The apterae of this species are readily recognized
by the black horseshoe-shaped marking on the dorsum of the abdomen.

89. Myzus cynosbati (Oestlund)

Oestlund, Minn. Geol. and Nat. Hist. Surv., Bull. 4, p. 81, 1887. Nectaro-

phora (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 10, p. 294, 1917 (note).
Shinji, Can. Ent., vol. 49, p. 49, 1917. M. ribis (Linn.) (list).

Records. — Eibes vulgar e; Walnut Creek (Davidson) ; Eibes glutinosum, E.
menziesii; Berkeley, April, 1915 (Shinji).

This species has been taken but a few times in the San Francisco
Bay region; once on cultivated red currant in company with Aphis
neomcxicana pacifica, once on wild flowering currant, and once on
wild canyon gooseberry. Furthermore, only the sexapura (migrants)
and sexuales have been taken. Davidson writes that this is true
cynosbati of Oestlund and not the species described by Davis (Ann.
Ent. Soc. Am., vol. 2, p. 38, 1909), as Macrosiphum cynosbati (Oest.),
which is not that species but some other. Shinji listed M. ribis
(Linn.), but his specimens prove to be the sexuales of this species.

90. Myzus fragaefolii Cockerell

Figure 177

Cockerell, Can. Ent., vol. 33, p. 101, 1901 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. X35, 1914 (desc. sexuales).

Records. — Fragaria chiloensis; Walnut Creek, Contra Costa County (David-
son); Berkeley, March to September, 1915; Palo Alto, April, 1915; Ontario,
April, 1917; Buena Park, Orange County, May, 1917 (E. K. Bishop); Santa
Barbara, May, 1917; Eialto, San Bernardino County, May, 1917 (A. B. Snow):
F. californicus; Pine Hills, San Diego County, June, 1916.

On the under side of the leaves of native and cultivated straw-
berries this small yellowish aphid is often found, both in the San
Francisco Bay region and in southern California. Seldom does it
become abundant, although several records of its abundance were
received from various parts of the south during the spring of 1917.
Several growers have thought it bad enough to spray for it. During
the late winter (January and February) the sexuales appear and
the eggs are laid. These hatch in a short time, and during the rest
of the year the alate and apterous viviparae are found.

76 MISCELLANEOUS STUDIES

91. Myzus lycopersici (Clarke)

Clarke, Can. Ent., vol. 35, p. 253, 1903. Nectaropliora (orig. desc.).
Davis, Can. Ent., vol. 46, p. 123, 1914 (desc.).

Eecord. — Lycopersicum esculentum ; Berkeley (Clarke).

Only once has this species been found in California. Davis in
1914 described a species from tomato in Idaho, Montana, and Oregon
which he believed to be this one. It may be, and it may not be so.
That can never be decided for the types of -Clarke's species are all lost.

92. Myzus rhamni (Clarke)

Figure 178

Clarke, Can Ent., vol. 35, p. 254, 1903. Nectaropliora (orig. desc.).
Shinji, Can. Ent., vol. 49, p. 49, 1917. M. rhamni (Boyer) (list).

Records. — Rhamnus calif ornicus; Berkeley (Clarke), Berkeley, March, 1915
(Shinji).

In March, 1915, George Shinji took a species of Myzus from coffee-
berry in Berkeley. This fits Clarke's description of Nectarophora
rhamni, in so far as the description goes. The author considers it to
be the same species as described by Clarke, inasmuch as it was collected
in the same locality and on the same host plant.

Wilson (Can. Ent., vol. 44, p. 156, 1912) describes a species from
Rhamnus purshiana in Oregon as M. rhamni (Boyer), listing Clarke's
species as a synonym. This is the same species as taken by Shinji
in Berkeley, but it is doubtful if it is the species described by Boyer
de Fonscolombe. Specimens in the author's collection from Rhamnus
in Colorado are determined by Gillette and Bragg to be Aphis rhamni
Fonsc. These are certainly different from the coast species, the former
being an Aphis closely related to A. euonomii Fabr., the latter a
Myzus. From this evidence the author cannot follow "Wilson in
placing Nectarophora rhamni Clarke as a synonym of Aphis rhamni
Fonsc., considering both as Myzus, but he considers them as distinct,
Clarke's species being a Myzus, Fonscolombe 's an Aphis.

93. Myzus ribifolii Davidson
Davidson, Jour. Econ. Ent., vol. 10, p. 294, 1917 (orig. desc.).

Record. — Ribes glutinosum; Eedwood Canyon, Contra Costa County (David-
son).

A SYNOPSIS OF THE APHIDIDAE 77

Davidson recently described all forms of this species from speci-
mens taken during March, April, and May, 1913, 1914, and 1915, on
wild flowering currant in Redwood Canyon, Contra Costa County.
The author is unacquainted with the species.

94. Myzus varians Davidson

Davidson, Jour. Econ. Ent., vol. 5, p. 409, 1912 (orig. desc.).
Record. — Clematis ligusticifolia ; San Jose (Davidson).

Davidson found this species on the under side of the leaves of
wild clematis, or Yerbade chivato, near San Jose, and later in Walnut
Creek, Contra Costa County. The author is unacquainted with the
species.

25. Genus Nectarosiphon Schouteden

Schouteden, Aphidologische Notizen, Leipzig, 1901. Type Macrosiphum
rubicola Oestlund, n.n. for Macrosiphum Oestlund, preoccupied.

KEY TO CALIFORNIA SPECIES
1. Body quite large, being about 3 to 4 mm. in length. Wings with dusky spot

near tip rubicola (Oestlund)

— Body not so large, being only about 1.5 mm. long. Wings without dusky spot
near tip morrisoni Swain

95. Nectarosiphon rubicola (Oestlund)

Figures 107, 109, 123

Oestlund, Minn. Geol. Nat. Hist. Surv., vol. 14, p. 27, 1886. Macrosiphum

(orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. 136, 1914. Amphorophora (list).

Eecords. — Hubus nutlcanus; Contra Costa County (Davidson) ; Berkeley (Essig,
Shinji).

This species is sometimes found infesting the tender leaves and
shoots of thimbleberry in the San Francisco Bay region. The most
distinctive character which readily separates it from Amphorophora
rubi (Kalt.) is the presence of a dusky patch near the tip of the fore-
wing. This was originally described by Oestlund as the type of his
genus Macrosiphum. However, this name was preoccupied by Macro-
siphum Passerini, so Schouteden proposed the name Nectarosiphon
for this genus. Davidson listed this species as Amphorophora, and
Morrison writes that he has never been able to satisfy himself why

78 MISCELLANEOUS STUDIES

this is not Amphorophora instead of Nectar osiphon. There is con-
siderable difference in the antennal tubercles of this species and
species of Amphorophora, although otherwise they are quite similar.
The author believes that slight as the difference is it should be recog-
nized for it is through the shape and size of the antennal tubercles
that the different genera of the Macrosiphini are recognized in a
large part. In this species the tubercles are large and distinct and
neither gibbous nor toothed on the inner side, and with the outer side
quite evident, while in Amphorophora they are small and distinctly
toothed on the inner side, with the outer side a mere line, or not at
all evident.

96. Nectarosiphon morrisoni Swain

Figures 124 to 127
Swain, Trans. Am. Ent. Soc., vol. 44, p. 8, 1918.

Records. — Cupressus macrocarpa; San Francisco (Compere, Morrison), San
Diego (Swain) : C. guadalupensis; San Diego (Swain).

In Golden Gate Park, San Francisco, and in Exposition Park, San
Diego, this species has been taken on cypress. The small, slender,
long-legged apterae are found infesting the terminal leaves of the
host. Occasionally an alate female is seen. In San Diego, the apterae
were found in company with Cerosipha cupressi Swain.

26. Genus Pentalonia Coquerel

Coquerel, Ann. Ent. Soe. France, vol. 7, p. 239, 1860. Type P. nigro-
nervosa n.sp.

97. Pentalonia nigronervosa Coquerel

Coquerel, Ann. Ent. Soe. France, vol. 7, p. 239, 1860 (orig. desc.).
Wilson, Jour. Econ. Ent., vol. 2, p. 346, 1909 (desc.).

Record. — Pelargonium sp. ; Stanford University (Morrison).

The following note concerning this species is from Morrison:
Pentalonia nigronervosa Coquerel. See Wilson, Jour. Econ. Ent., 1909. In
the Davidson collection (belonging to Stanford University) there is a single
glycerine jelly mount of this species. I have been able to see enough of it to be
certain of its identity with that described by Wilson in the Journal (above). The
record is from geranium, and Davidson once told me that he found it in alcohol
in the laboratory [of Stanford University] at the time he began his study of the
Aphididae. I believe the record should be published.

A SYNOPSIS OF THE APHIDIDAE 79

27. Genus Phorodon Passerini
Passerini, Gli Afidi, 1860. Type P. humuli Schr.

No attempt has been made to formulate a key to the California
species of this genus, owing to the fact that the author has specimens
of but one species, and that the description of the other is quite inade-
quate. Four species have been reported from this state, two of which
prove to be species of other genera and one of which is very doubtful.
Phorodon carduinum (Walker) as reported by Davidson, is Myzus
braggi Gillette. Phorodon galeopsidis (Kaltenbach), also reported by
Davidson, is Rhopalosiphum hippophoaes Koch. There is much
diversity of opinion concerning the specific determination of these
species and of Myzus elaeagni "Del Guercio. One might refer to Gil-
lette 's paper on Rhopalosiphum hippophaaes Koch and Myzus braggii
Gillette. Davis writes that he is not prepared to be quoted. Davidson
lists P. galeopsidis and R. hippophoaes as synonyms. He states that
his specimens listed as P. carduinum Walker were determined by
Monell, but that Van der Goot is doubtful, while he himself believes
them to be M. braggii Gillette. He has been followed in so listing
them. This then leaves but two species reported from California.

98. Phorodon humuli (Schrank)
Figures 115 to 118

Schrank, Fauna Boica, vol. 2, p. 110, 1801-02. Aphis (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 252, 1903 (list).

Clarke, Calif. Agri. Exp. Sta., Bull. 160, 1904 (econ.).

Parker, U. S. Dept. Agri., Bull. Ill, 1913 (econ.).

Vosler, Mon. Bull., Cal. Comm. Hort., vol. 2, p. 668, 1913 (list).

Records. — Humulus spp. ; Berkeley (Clarke); Placer County' (Vosler) ; Berke-
ley, July to September, 1915: Prunus domestica; Berkeley, March to April, 1915
(Essig, Shinji) ; (Parker).

This is the common hop plant louse found throughout the central
part of the state. During the summer it is common on hops, but in
the fall the sexupara migrate to plum, where the eggs are laid. These
eggs hatch the following spring into stem mothers which feed on the
opening buds of plum. During later generations, probably about the
third or fourth, alate fundatrigeniae appear, which leave the plum
and migrate to hop. Here the summer generations are produced until
well into the fall. Parker states that the normal life cycle is as just
stated, but that it is also possible, and it occasionally occurs, that this

80 MISCELLANEOUS STUDIES

aphid may live the entire year upon hops, or on plum, generation after
generation of parthenogenetic females being produced.

99. Phorodon scrophulariae Thomas

Thomas, Ann. Eep. 111. St. Ent., vol. 8, p. 72, 1879 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 252, 1903 (list).

Eecord. — Scrophularia sp., Berkeley (Clarke).

This is a doubtful species, reported by Clarke as present on
Scrophularia in Berkeley, and by Dr. Thomas in 1879 on a species
of plant which he thought to be Scrophularia in Illinois. Since
Clarke's record it has never been found, although Morrison states
that he has spent considerable time examining the common Scrophu-
laria plants in the vicinity of Stanford University, but to no avail.
The author attempted to find it many times in the vicinity of San
Diego during 1916, and in the vicinity of Riverside in 1917, with no
success.

28. Genus Rhopalosiphum Koch
Koeh, Die Pflanzenlause, p. 23, 1854. Type Aphis persicae Sulz.

This genus is very closely related to Myzus, and is distinguished
only by the shape of the cornicles. This distinction is variable, how-
ever, as in some species certain forms have the clavate cornicles of
Rhopalosiphum while other forms have the cylindrical cornicles of
Myzus. This is particularly true in the case of Rhopalosiphum
persicae (Sulz.) and Myzus braggii Gillette. However, most aphidol-
ogists separate these two genera, so the author feels that it is best
to do so.

KEY TO CALIFORNIA SPECIES

Alate viviparous females

1. Ground color dark (olive-green, wine, brown, and so forth) 2

— Ground color light, usually green (this does not refer to the dark markings

on head, thorax, or abdomen, but rather to the ground color of the
abdomen) 4

2. Wing veins with smoky borders and tips (fig. 164). IV with a few small

sensoria violae Pergande

— Wing veins without smoky borders or tips, and IV without sensoria.

rhois Monell

3. Antennae distinctly tuberculate, with sensoria on both III and IV (figs. 170,

279) 4

— Antennae not tuberculate, and IV without sensoria, or at most with but a

few small ones (figs. 167, 168) 5

A SYNOPSIS OF TEE APEID1DAE 81

4. VI spur slightly longer than III (figs. 279, 281). Cornicles quite large and

heavy (figs. 282, 284) lactucae (Kalt.)

— VI spur about twice as long as III. Cornicles comparatively small and slender

(fig. 165 ) hippophoaes Koch.

5. First discoidal vein with distinct, smoky border, second discoidal bordered

slightly so (fig. 166) nervatum Gillette

— First and second discoidal without smoky borders 6

6. Abdomen with dusky dorsal markings. Ill with a few (10-12) sensoria

(fig. 168) persicae Sulz.

— Abdomen without dusky dorsal markings. Ill with many (24-30) sensoria

(fig. 167) corylinum Davidson

Apterous viviparous females*2
1. Ground color dark (olive-green, wine, brown) 2

— Ground color light (green, and so forth) 3

2. Cornicles large and stout, longer than III rhois Monell

— Cornicles smaller and more slender, shorter than III violae Pergande

3. Cornicles longer than III 4

— Cornicles shorter than III 5

4. VI spur considerably longer than III, and subequal to cornicles.

nervatum Gillette

— VI spur about equal to III, and distinctly shorter than cornicles.

hippophoaes Koch.

5. Ill with secondary sensoria lactucae (Kalt.)

— Ill with no secondary sensoria persicae (Sulz.)

100. Rhopalosiphum corylinum Davidson

Figure 167

Davidson, Jour. Econ. Ent., vol. 7, p. 134, 1914 (orig. desc.).
Records. — Corylus rostrata; Walnut Creek, Contra Costa County (Davidson) :
Physocarpus capitatus; (Davidson).

This species was originally described from specimens of alate
viviparae and pupae taken on wild hazelnut near Walnut Creek.
Davidson writes that he has found it quite common on nincbark in
the San Francisco Bay region. The author has never taken the
species, but has had access to cotype specimens in Essig's collection.

101. Rhopalosiphum hippophoaes Koch

Figures 165, 170

Koch, Die Pflanzenlause, p. 28, 1854 (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 7, p. 136, 1914. Phorodon galeopsidis

Kalt. (list).
Gillette, Jour. Econ. Ent., vol. 8, p. 375, 1915 (synonomy).

Record. — Polygonum sp.; San Jose (Davidson).

12 R. corylinum Dvdn. is omitted from this key as the apterous female was
never described and specimens are not available to the author.

82 MISCELLANEOUS STUDIES

Davidson reported this species as present on knotweed in the
vicinity of San Jose, under the name P. galeopsidis Kalt. Later he
followed Gillette in placing it as a synonym of R. hippophoaes Koch.
The author has never collected it, but has had access to specimens
from Davidson in San Jose, and Davis in Oak Park, Illinois. For a
full discussion of the synonymy of this species see Gillette's paper
listed above.

102. Rhopalosiphum lactucae (Kalt.)

Figures 277 to 285

Kaltenbach, Monog. d. Pflanzenlause, p. 37, 1843. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 277, 1910 (list).

Records. — Sonchus spp. ; Stanford University (Davidson); Stanford Univer-
sity, May to July, 1915; Walnut Creek, May, 1915 (Davidson); Berkeley, July,
1915; Lemon Grove, San Diego County, January, 1916; Biverside, January to
May, 1917; Los Angeles, April, 1917: Asclepias sp.; Corvallis, Oregon, November,
1913 (Moznette).

This is a common species infesting the heads of sow thistle
throughout the San Francisco Bay region and southern California.
In November, 1913, G. F. Moznette took it on milkweed in Corvallis,
Oregon. This collection consisted entirely of alate females, that may
have been the sexupara. Inasmuch as the identity of this species is
doubtful there is given below a brief description drawn from speci-
mens of nine alates and eight apterae taken on Sonchus spp. at Stan-
ford University in May, 1915, in Walnut Creek in May, 1915, in
Berkeley in July, 1915, and in Lemon Grove in January, 1916, and
on Asclepias sp. in Corvallis, Oregon, in November, 1913. This latter
collection is by George F. Moznette of Corvallis.

Alate viviparous female. — Prevailing color is apple green with
the head dark green to black, the prothorax apple green, the thoracic
lobes black. The abdomen is apple green with three pair of dusky
marginal spots on segments one, two, and three, respectively, and
with a larger dusky patch on the dorsum of segments four, five, and
six, being between the cornicles. The cornicles and cauda are luteous
with the extreme tip of the former dusky. The antennae are dusky
throughout. The legs are luteous with the tarsi and tips of the femora
and tibiae dusky.

The head is about twice as broad as long, with a distinct frontal
tubercle (fig. 278). The antennae are set on distinct tubercles and
are between one and one-fourth to one and one-half times as long

A SYNOPSIS OF THE APHIDIDAE 83

as the body. The relative lengths of the segments are as follows: the
spur is the longest, being followed by III, which is subequal but never
longer. IV is about one-half the length of the spur and slightly longer
than V. II is slightly longer than VI, which is about equal to I.
Sensoria are arranged as follows (figs. 279-281) : on V and VI are
the usual primary and accessory sensoria; on V in addition to the
primary sensoria, there are at times as many as seven small circular
secondary sensoria, located about the middle of the segment. The
number of these sensoria range from none to seven, two and three
being the usual number ; on IV there are from six to twelve irregular
secondary sensoria (fig. 280), placed irregularly along the whole
length of the segment ; on III there are between thirty and forty
irregularly placed and irregularly sized sensoria (fig. 279) scattered
along the whole length of the segment. The usual number is from
thirty-six to thirty-nine. The prothorax is without lateral tubercles.
The beak is of medium length, reaching to slightly beyond the second
coxae. The cornicles (fig. 282) are fairly large and clavate on one
side. At the widest point they are slightly less than one-fifth the
length. The tip is slightly wider than the base. They are about the
same length as the fourth antennal segment, although in some cases
they may be slightly longer, and in others slightly shorter, but in all
cases longer than the fifth antennal segment. The cauda (fig. 283) is
long and fairly large, not quite reaching to the tip of the cornicles,
being about one-half as long as the cornicles and one-half as long
again as the hind tarsi. The- wings and venation are normal, the
forewings being about twice as long as the body.

Measurements : Body length, 1.48 to 1.87 mm. ; width, 0.73 to 0.82
mm. ; antennae total, 2.35 to 2.51 mm. ; III, 0.544 to 0.697 mm. ; IV,
0.306 to 0.425 mm. ; V, 0.218 to 0.357 mm. ; VI, 0.085 to 0.119 mm. ;
spur, 0.68 to 0.799 mm. ; cornicles, 0.323 to 0.459 mm. ; cauda, 0.187 to
0.255 mm. ; hind tarsi, 0.136 to 0.153 mm. ; wing length, 3.4 to 3.8 mm. ;
wing width, 1.2 to 1.5 mm. ; wing expansion, 8.0 to 8.3 mm. The
average measurements are as follows : body length, 1.74 mm. ; width,
0.768 mm. ; antennae total, 2.445 mm. ; III, 0.645 mm. ; IV, 0.382 mm. ;
V, 0.328 mm. ; VI, 0.107 mm. ; spur, 0.753 mm. ; cornicles, 0.403 mm. ;
cauda, 0.248 mm. ; hind tarsi, 0.139 mm. ; wing length, 3.6 mm. ; width,
1.32 mm. ; expansion, 8.1.

Apterous viviparous female. — Prevailing color pale green with the
head paler, being almost luteous or of a pale yellowish green color.
The eyes are red. The antennae, except the apices of segments three

84 MISCELLANEOUS STUDIES

to six inclusive, the legs, except the tarsi and tips of the tibiae, the
cauda, and the cornicles, except the tip, are all luteous. Sensoria are as
follows : on V and VI the usual primary sensoria, on VI the accessory
sensoria, and on III (fig. 278), from nine to eleven small, circular
irregularly placed secondary sensoria. IV is without sensoria. The
antennae are considerably longer than the body, the spur and III
being subequal and the longest segments. Sometimes the spur is
slightly longer than III. V is about one-half as long as III or the
spur, and about four-fifths as long as IV. I and VI are subequal,
being about one-seventh as long as the spur. The cornicles (fig. 284),
are clavate, quite large, usually being slightly more than one-fifth
the length of the body and over three times the length of the hind
tarsi. The cauda (fig. 285) is long, sickle-shaped, and a little more
than one-half as long as the cornicles.

Measurements : Body length, 1.7 to 2.18 mm. ; width of abdomen,
0.82 to 1.73 mm.; antennae total, 2.32 to 2.48 mm.; Ill, 0.646 to
0.714 mm. ; IV, 0.391 to 0.425 mm. ; V, 0.323 to 0.34 mm. ; VI, 0.102
mm. ; spur, 0.646 to 0.782 mm. ; cornicles, 0.459 to 0.493 mm. ; cauda,
0.238 to 0.272 mm.; hind tarsi, 0.136 to 0.153 mm. The average
measurements are as follows : body length, 1.87 mm. ; width, 0.99 mm. ;
antennae total, 2.39 mm. ; III, 0.674 mm. ; IV, 0.408 mm. ; V, 0.334
mm. ; VI, 0.102 mm. ; spur, 0.7099 mm. ; cornicles, 0.473 mm. ; cauda,
0.255 mm. ; hind tarsi, 0.1445 mm.

103. Rhopalosiphum nervatum Gillette

Figures 166, 169, 171

Gillette, Can. Ent., vol. 40, p. 63, 1908 (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 3, p. 378, 1910. B. arbuti, n.sp. (desc.).

Davidson, Jour. Econ. Ent., vol. 7, p. 134, 1914 (list).

Records. — Arbutus menziesii; Stanford University, San Jose, Walnut Creek
(Davidson); Sacramento (Essig) ; Stanford University, February to May, 1915;
Berkeley, September, 1915: Arbutus unedo; Eedlands, February, 1917; Bosa spp. ;
Walnut Creek (Davidson) ; Berkeley, February, 1915 (Essig).

In 1910 Davidson described a species of Rhopalosiphum, which he
named arbuti, from specimens taken on madrone in the vicinity of
Stanford University. Since then it has been found quite commonly
on madrone throughout the San Francisco Bay region, and once on
a strawberry tree in Silva Park, Redlands. It was noticed by the
author that the alate females were very scarce at all times, although
the apterae and nymphs were often quite abundant. Later, when

A SYNOPSIS OF THE APE I DID AE 85

studying specimens while working up a key to the species of Rhopalo-
siphum, he found that structurally this species was identical with
Rhopalosiphum nervatum Gillette. The latter had been taken on roses
in the San Francisco Bay region. The identical structure and the
scarcity of alates on madrone led to a belief that they were the same
species. However, it was too late in the season (October, 1915) to
try any transfer tests. No opportunity was found to try migration
tests until in February, 1917, when the species was taken in Redlands.
Two alate females were reared in the laboratory and then placed
under a muslin bag on a rose bush, out of doors. A few days later
these were examined and several young larvae observed. No further
observations were made for two weeks, when it was found that the
bag had been ripped off by the severe winds. Although this test was
not a complete success the author feels confident of the identity of
this species.

104. Rhopalosiphum persicae13 (Sulz.)

Figures 108, 119, 120, 168

Sulzer, Kan. Ins., p. 105, 1761. Aplite (orig. dese.).

Clarke, Can. Ent., vol. 35, p. 252, 1903. Rhopalosiphum dianthi (Schrank)

(list).

Gillette, Jour. Econ. Ent., vol. 1, p. 359, 1908. Myzus (desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 303, 1909. B. dianthi (Schrank),

E. achyrantes Monell, and Myzus (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. E. tulipae Thomas (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 378, 1910. E. dianthi (Schr.) (list).
Davidson, Jour Econ. Ent., vol. 3, p. 379, 1910. Mysus (list).
Essig, Pom. Jour. Ent., vol. 3, p. 598, 1911. Myzus (desc.).

Eecords. — Throughout California by Clarke, Davidson, Essig, Ferris, Morrison,
and the author on Abutilon sp., Amaranthus retro fiexus, Amsinclcia respectabilis,
Bougainvillaea sp., Brassica spp., Capsella bursa-pastoris, Capsicum annuum,
Catalpa sp., Chcnopodium murale, Citrus spp., Cynoglossum grande, Ci/ticus pro-
liferus, Geranium carolinianum, Hedera helix, Lycopersicum esculentum, Malva
parviflorus, Oxalis oregona, Prunus spp., Eanunculus califomicus, Eaphanus
sativus, Eumex spp., Sambucus glauca, Sanicula menziesii, Senecio vulgare,
Solanum tuberosum, Sonchus spp., Tropaeolum sp., Tulipa sp., Vinca major.

This green peach aphis is one of the most common aphids found
in the state. It is most abundant in the spring, at which time it will
be found on almost any plant. According to Gillette various species

is George Shinji (Can. Ent., vol. 49, p. 49, 1917) recently described an aphid
from specimens taken on Godetia amaena in Berkeley, which he named Myzus
godetiae n.sp. The author has not seen specimens of this species, but from the
description and figures, it is in all probability Ehopalosiphum persicae (Sulz.).

8t> MISCELLANEOUS STUDIES

of Prunus are the winter hosts in Colorado, while during the summer
it migrates to other plants. In California, however, winter eggs are
not laid, the viviparous females living the year round. So far as the
author has observed in over three years, only the form with clavate
cornicles is found in California.

105. Rhopalosiphum rhois Monell

Figure 173

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 27, 1879 (orig. desc.).
Davis, Can. Ent., vol. 46, p. 165, 1914. E. Iwwardi (Wils.) (desc.).
Essig, Univ. Calif. Publ. Ent., vol. 1, p. 330, 1917. B. howardi (Wils.)

(list).
Ibid., p. 334, 1917 (list).

Eecords. — Ehus diversiloba; Berkeley, April, 1915; Avena sativa, Berkeley,
(Essig).

This species has been taken in Berkeley on poison oak and grasses.
Essig reported it recently as R. howardi (Wils.), but according to
Gillette14 this is a synonym of R. rhois Monell, Rhus being the winter
host, and various species of GramMiaceae the summer hosts.

This species does not seem to be a typical Rhopalosiphum, being
quite close to Siphocoryne nymphaeae Linn., but yet not fitting the
generic description of Siphocoryne exactly. Consequently it is best
to list it as has been done heretofore as Rhopalosiphum. .

106. Rhopalosiphum violae Pergande
Figures 164, 174

Pergande, Can. Ent., vol. 32, p. 29, 1900 (orig. desc.).
Essig, Pom. Jour. Ent., vol. 1, p. 4, 1909 (desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 303, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 277, 1910 (list).

Eecords. — Viola spp. ; Claremont, Santa Paula (Essig); Stanford University
(Davidson); Palo Alto, May, 1915; Santa Ana, February, 1917; Eiverside, April,
1917.

This beautiful little aphid is found more or less abundantly in the
spring on the under side of the leaves of violets throughout the state.
The dark red color and broad black wing veins serve to distinguish it
readily from other aphids.

Gillette, Jour. Econ. Ent., vol. 8, p. 100, 1915.

A SYNOPSIS OF TEE APHIDIDAE 87

Tribe Aphidini Wilson

Wilson, Ann. Ent. Soc. Am., vol., 3, p. 331, 1910.
Following is a brief characterization of this tribe, from Wilson:

The characters which separate this tribe from the previous one [Macrosiphini]
are taken as follows: Antennae shorter than the body, or when as long as the
body the cornicles and cauda are very short; antennal tubercles, when present,
are indistinct, or else the cornicles and cauda are small; when the cornicles are
very long or large the development is limited and the other characters are used
to place the genera.

The California genera included by Wilson in this tribe are Aphis,
Cerosipha, Coloradoa, Hyalopterus, Liosomaphis, and Siphocoryne
[Hyadaphis], In addition to these the author includes Toxoptera
because of the small and indistinct antennal tubercles and the short
cornicles, and Myzaphis because of the absence of antennal tubercles.
The key to the California genera has been formulated by the author,
following Wilson, Mordwilko, and Van der Goot.

1. Antennae five-segmented Cerosipha del Guercio

— Antennae six-segmented 2

2. Cornicles much shorter than cauda Hyalopterus Koch

— Cornicles about as long as or longer than cauda 3

3. Cornicles cylindrical, tapering, or conical, not distinctly clavate (fig. 182),

except in Coloradoa and Myzapliis, in which they may be slightly clavate
at the apex (fig. 315) 5

— Cornicles distinctly clavate (figs. 183, 184) 4

4. Cornicles long and strongly clavate on one side (fig. 184). Antennae shorter

than body, with VI spur not longer than III Liosomaphis Walker

— Cornicles slender and but slightly clavate (fig. 183). Antennae never much

shorter than body, with VI spur longer than III (fig. 258) or with a supra-
caudal tubercle (figs. 255, 256) Siphorcoryne Passerini

5. Third discoidal vein but one-branched (fig. 276). Body without lateral

tubercles. Cauda long and prominent, being about as long as cornicles.

Toxoptera Koch

— Third discoidal vein twice-branched. Body with or without lateral tubercles.

Cauda usually distinctly shorter than cornicles 6

6. Front of head with a very distinct tubercle (figs. 308, 313). Body long

without lateral tubercles. Cornicles long and often slightly swollen near
apex Myzaphis Walker and Coloradoa Wilson

7. Front of head without prominent tubercle (fig. 233). Body more rounded

with lateral tubercles on prothorax and seventh abdominal segment, and
oftentimes on some of the anterior abdominal segments Aphis Linn.

88 MISCELLANEOUS STUDIES

29. Genus Aphis Linn.

Linnaeus, Syst. Nat., 1748. Type Aphis rumicis Linn.
KEY TO CALIFORNIA SPECIES
Alate viviparous females

1. Abdomen with floceulent masses of wax. Antennae considerably shorter than

body, and VI spur shorter than III alamedensis Clarke

— Abdomen without such floceulent masses of wax (except perhaps Aphis cooki

Essig) _ 2

2. Antennae one and one-half times as long as the body, or more.

houghtonensis Throop

— Antennae not so much longer than body; when longer, which is seldom, but

slightly so 3

3. Abdomen pale yellowish green. Found only on Moms sp inori Clarke

— Abdomen darker being black, dark green, yellow. Not found on Moms sp. 4

4. Abdomen dark-green with an orange band between the cornicles.

augelicae Koch

— Abdomen without such an orange band between the cornicles (sometimes there

is a slight orange or reddish coloring between the cornicles of the apterae
of Aphis avenae Fabr., but it is not constant) 5

5. Abdomen sage-green with faint lateral spots. Ill with apical one-half con-

spicuously darkened and with six large sensoria. VI spur less than one-half
as long as III. On Atriplex spp tetrapteralis Cockerell

— Not with above combination of characters 6

6. IV with secondary sensoria (fig. 244) 7

— IV without secondary sensoria (fig. 204) 27

7. Cornicles and hind tarsi subequal 8

— Cornicles considerably longer than hind tarsi 17

8. VI spur shorter than III 9

— VI spur equal to or longer than III '. 13

9. Cornicles short and tapering 10

— Cornicles short and incrassate pseudobrassicae Davis

10. V with secondary sensoria. Body slightly pulverulent cooki Essig

— V without secondary sensoria. Body not pulverulent 11

11. Less than 12 secondary sensoria on III, arranged in a more or less even line 12

— About 20 to 25 secondary sensoria on III, arranged irregularly along segment

(fig. 244) senecio Swain

12. Ill with 9 to 12 sensoria. V and VI base subequal, each being shorter than

IV lithospermi Wilson

— Ill with 5 to 9 sensoria. IV and V subequal, each being longer than VI

base viburnicolens n.sp.

13. Cornicles shorter than hind tarsi. A large black species in life being marked

with white bars and cross bands on the abdomen albipes Oestlund

— Cornicles and hind tarsi subequal. Body color greenish 14

14. Root-infesting species. Antennae short, scarcely reaching the middle of the

abdomen mlddletonii Thomas

— Aerial species. Antennae reaching to base of the cornicles, or as long as

body 15

15. Cornicles incrassate. A medium-sized species pseudobrassicae Davis

— Cornicles cylindrical and tapering slightly. A smaller-sized species 16

A SYNOPSIS OF THE APHIDIDAE 89

16. Cauda shorter than hind tarsi. Ill with 11 to 15 sensoria scattered irregu-

larly along segment (fig. 294) marutae Oestlund

— Cauda longer than hind tarsi. Ill with 5 to 9 more or less evenly arranged

sensoria viburnicolens n.sp.

17. Cornicles equal to or longer than III 18

— Cornicles not as long as III 19

18. Cauda, cornicles, and III subequal. Second branch of third discoidal vein

very near to apex of wing spiraecola Patch

— Cauda considerably shorter than cornicles or III, the last two being subequal.

Second branch of third discoidal about midway between base of first
branch and apex of wing oenotherae Oestlund

19. Fore wing with the second branch of the third discoidal arising very near

to the apex of the wing. (In a few cases the second branch is not found,
but never in both wings) (fig. 191) avenae Fabr.

— Venation of fore wing normal (fig. 187) 20

20. Antennae longer than body persicae-niger Smith

— Antennae not longer than body 20

21. A pair of small tubercles present on the middle of the seventh and eighth

abdominal segments malifoliae Fitch

— Such tubercles not present 22

22. V with secondary sensoria. VI spur longer than III 23

— V without secondary sensoria. VI spur at most equal to III 25

23. Beak scarcely reaching second coxae maidis Fitch

— Beak reaching beyond second coxae, even to or beyond the third 24

24. Cornicles longer than cauda (figs. 194, 195) and more than twice as long as

hind tarsi sambucifoliae Fitch

— Cornicles and cauda subequal; the former not more than twice as long as

hind tarsi neomexicana Cockerell var. pacifica Davidson

25. Cauda and hind tarsi subequal. Ill with a few large sensoria (fig. 232).

Abdomen green with dark dorsal markings ramona Swain

— Cauda longer than hind tarsi. Ill with several sensoria. Abdomen black

or dark brown 26

26. Cornicles more than twice as long as hind tarsi, often almost three times as
long. VI spur and cornicles subequal, hind tarsi and. VI base subequal.

hederae Kalt.

— Cornicles never more than twice as long as hind tarsi, usually considerably

less. Hind tarsi usually slightly longer than VI base, and VI spur longer
than cornicles euonomi Fabr.

27. Cornicles distinctly knobbed, the tip being widened to twice the width of

the rest of the cornicles frlgidae Oestlund

— Cornicles normal 28

28. Fore wing with the second branch of the third discoidal arising very near the

apex of the wing (fig. 188) salicicola Thomas

— Fore wing with venation normal (fig. 187) 29

29. Cornicles distinctly longer than cauda 31

— Cornicles at most equal to cauda 30

30. Cornicles short and swollen throughout apical one-half (fig. 203). Antennae

as long as or longer than the body .„ brassicae Linn.

— Cornicles short and slender, and slightly clavate on one side. Antennae

scarcely two-thirds as long as the body atriplicis Linn.

31. Abdomen without lateral tubercles on anterior segments. Cauda short and

broad, with rounded tip, and almost as long as the cornicles cardui Linn.

— Abdomen with lateral tubercles on at least one of the anterior segments .... 32

90 MISCELLANEOUS STUDIES

32. VI spur shorter than III 33

— VI spur not shorter than III 34

33. Cornicles about three times as long as cauda medlcaginis Koch

— Cornicles not three times as long as cauda 34

34. Cauda more than one-half as long as cornicles 35

— Cauda not more than one-half as long as cornicles 38

35. Ill with four or five fairly large semsoria oregonensis Wilson

— Ill with many irregular sensoria 36

36. Ill with 20 or more sensoria, IV with none 37

— Ill with less than 20 sensoria, usually 14 or 15. IV usually with one or two,

or. more sensoria euonomi Fabr.

37. IV about one-third longer than V. Cornicles about four times as long as

broad at base. On Heraclium spp heraclii Cowen

— IV but about one-sixth longer than V. Cornicles about three times as long

as broad at base. On Yucca sp Yuccae Cowen

38. A few (about 10) equal-sized sensoria on III (fig. 222). A large yellow

species with distinct dark markings nerii Fonsc.

— About 20 irregular sensoria on III (fig. 211). Not yellow 39

39. Cornicles slightly more than twice as long as hind tarsi 40

— Cornicles not twice as long as hind tarsi carl Essig

40. Hind tarsi slightly longer than cauda ceanothi Clarke

— Hind tarsi shorter than cauda cornifoliae Fitch

41. VI spur one and one-half or more times as long as III setariae Thomas

— VI spur never so much longer than III 42

42. Ill with a few large circular sensoria (5-10) (figs. 226, 290) _ 43

— Ill with several (15 or more) irregular sensoria 45

43. Beak reaching to or beyond third coxae. IV never with sensoria.

gossypii Glover

— Beak not reaching third coxae 44

44. VI spur longer than III (fig. 226). Small size pomi de Geer

— VI spur subequal to or shorter than III (figs. 289, 290). Medium to large

size cerasifoliae Fitch

45. Cornicles twice as long as cauda. Femora of all three pairs of legs similarly

colored carl Essig

— Cornicles longer than cauda, but not twice as loag. Femora of first pair of

legs pale, of second and third pair black euonomi Fabr.

Apterous viviparous females^

1. Cornicles shorter than hind tarsi 2

— Cornicles equal to or longer than hind tarsi 4

2. VI spur longer than III. White bars and bands on abdomen in life.

albipes Oestlund

— VI spur not longer than III. Abdomen not as above 3

3. Cornicles and cauda subequal. Beak not reaching to second coxae. Pul-

verulent brassicae Linn.

— Cornicles shorter than cauda. Beak reaching to or beyond second coxae. Not

pulverulent atriplicis Linn.

is In this key only those species are included of which there are specimens in
the author's collection or of which there are adequate descriptions available.
The following species are therefore omitted: Aphis alamedensis Clarke, A. hough-
tonensis Throop, A. mori Clarke, A. neomexicana Cockerell, A. oenotherae Oest-
lund, and A. tetrapteralis Cockerell.

A SYNOPSIS OF THE APHIDIDAE 91

4. Cornicles and hind tarsi subequal 5

— Cornicles longer than hind tarsi „ 10

5. Secondary sensoria on III and IV. Boot species _ middletonii Thomas

— No secondary sensoria. Aerial species 6

6. Ill longer than VI spur 7

— Ill shorter than or at most equal to VI spur 8

7. IV and cornicles subequal Jithospermi Wilson

— IV shorter than cornicles. Pulverulent cook! Essig

8. IV and cornicles subequal. Antennae considerably more than one-half the

length of the body 9

— IV shorter than cornicles. Antennae at most one-half the length of the body.

senecio Swain

9. Cornicles twice as long as cauda and slightly swollen before the tip.

avenae Fabr.

— Cornicles not twice as long as cauda, cylindrical, and tapering toward tip.

marutae Oestlund

10. Cornicles less than twice as long as hind tarsi 11

— Cornicles twice as long as or longer than hind tarsi 21

11. Secondary sensoria on III and IV. Eoot-infesting species.

middletonii Thomas

— No secondary sensoria. Aerial species 12

12. VI spur longer than III „ 13

— VI spur at most equal to III 16

13. Ground color, black or dark brown 14

— Ground color, a shade of green 15

14. VI spur one and one-half to two times as long as III. Apex only of femora

dusky setariae Thomas

— VI spur but slightly longer than III. Apical one-half of femora dusky.

medecaginis Koch

15. Pale green. Cornicles and cauda subequal. Dark, mottled green. Cornicles

twice as long as cauda or longer avenae Fabr.

16. VI spur considerably shorter than III 17

— VI spur almost as long as III 18

17. Cornicles swollen toward tip pseudobrassicae Davis

— Cornicles cylindrical and tapering toward tip ramona Swain

18. Cornicles but slightly longer than hind tarsi 19

— Cornicles about one and one-half times as long as hind tarsi 20

19. Dark green. Cornicles at least three times as long as broad at base.

maidls Fitch

— Pale green. Cornicles at most twice as long as broad at base.

senecio Swain

20. Dark green to reddish yellow. On Yucca spp yuccae Cowen

— Black or very dark brown with black dorsal bands and spots. On various

plants euonomi Fabr.

21. Cornicles distinctly knobbed at tip frigidae Oestlund

— Cornicles normal 22

22. VI spur longer than II '. 23

— VI spur at most equal to III 25

23. Pale green with dusky dorsal abdominal markings calendulicola Monell

— Not colored as above, either not green, or if green with dusky dorsal abdom-

inal markings 24

92 MISCELLANEOUS STUDIES

24. Bright yellow with black markings. Cornicles at least three times as long as

hind tarsi nerii Fonsc.

— Dark green with black markings. Cornicles but about twice as long as hind

tarsi cardui Linn.

25. Cornicles longer than III 26

— Cornicles at most equal to III 29

26. Ill considerably longer than VI spur 27

— Ill subequal to or but slightly longer than VI spur 28

27. Black. Cornicles about three times as long as hind tarsi. Ill one and one-

half times as long as VI spur sambucif oliae Fitch

— Green, pale to apple. Cornicles about four times as long as hind tarsi. Ill

almost twice as long as VI spur salicicola Thomas

28. Cornicles subequal to or but slightly longer than III, and about twice as

long as cauda prunorum Fabr.

— Cornicles one and one-half to two times as long as III, and about four times

as long as cauda oregonensis Wilson

29. Cornicles considerably shorter than III 30

— Cornicles subequal to or but slightly shorter than III 37

30. Ill and IV spur subequal persicae-niger Smith

- Ill longer than VI 1 31

31. Cornicles at least twice as long as cauda 36

— Cornicles not twice as long as cauda 32

32. Pale green, pulverulent cerasifoliae Fitch

— Dark green, brown, or black, not pulverulent 33

33. Cornicles about three times as long as hind tarsi 34

— Cornicles not three times as long as hind tarsi 35

34. Ill with a few small secondary sensoria hederae Kalt.

— No secondary sensoria .- cornifoliae Fitch

35. Cornicles considerably more than twice as long as hind tarsi. Lateral abdom-

inal tubercles only on first and seventh segments heraclii Cowen

— Cornicles at most but slightly more than twice as long as hind tarsi. Lateral

tubercles usually on more than first and seventh segments .... euonomi Fabr.

36. Antennae about as long as body. Cornicles more than twice as long as

Cauda cari Essig

— Antennae but about one-half as long as body. Cornicles but about twice as

long as cauda gossypii Glover

37. Ill considerably longer than VI spur : 38

— Ill and VI spur subequal 40

38. A pair of dorsal abdominal tubercles on sixth and seventh segments.

malifoliae Fitch

— No dorsal abdominal tubercles on sixth and seventh segments 39

39. Cornicles green, cylindrical, tapering slightly toward tip, and fairly straight.

Cauda about one and one-half times as long as hind tarsi. Abdomen with-
out dusky dorsal markings ramona Swain

— Cornicles black, cylindrical, curved outward. Cauda and hind tarsi subequal.

Abdomen with dusky dorsal markings seanothi Clarke

40. VI spur slightly longer than III. Cornicles and cauda subequal.

viburnicolens n.sp.

— VI spur slightly shorter than III. Cornicles one and one-half times as long

as cauda pomi De Geer

A SYNOPSIS OF THE APHIDIDAE 93

107. Aphis alamedensis Clarke

Clarke, Can. Ent., vol. 35, p. 251, 1903 (orig. desc.).
Record. — Prunus domestica; Berkeley (Clarke).

This is an unknown species described from specimens taken by
Clarke on greengage plum in Berkeley. Davidson suggests that it
might be Aphis cardui Lirin. (pruni Koch) from its brief description.

108. Aphis albipes Oestlund
Figures 198 to 200

Oestlund, Geol. Nat. Hist. Surv. Minn., Bull. 4, p. 52, 1887 (orig. dese.).
Williams, Univ. Neb. Studies, vol. 10, p. 119, 1910 (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).

Records. — Symphoricarpus racemosus; Stanford University (Davidson) ; Con-
gress Springs, Santa Clara County, July, 1915 (McCraeken) ; Berkeley, July, 1915
(Shinji).

This species is found at times curling the leaves of snowberry in
the San Francisco Bay region. Dr. McCraeken noted in connection
with the infestation at Congress Springs, "they are quite prettily
patterned with white bars and cross-bars." This is usually enough
to distinguish them.

109. Aphis angelicae Koch.

Koch, Die Pflanzenlause, p. 521, 1854 (orig. desc.).
Wilson, Jour. Econ. Ent., vol. 2, p. 348, 1909 (desc.).

Record. — Angelica sp., Hedera sp. ; California (Wilson).

Wilson reported this species from California, but gave no locality
or date. It is unknown to the author.

110. Aphis atriplicis Linn.

Linnaeus, Fauna Sweden, p. 1000, 1761 (orig. desc.).
Hayhurst, Ann. Ent. Soe. Am., vol. 2, pp. 88-100, 1909 (desc.).
Davidson, Jour. Econ. Ent., vol. 5, p. 407, 1912 (desc. sexuales apterous

viviparae).
Davidson, Jour. Econ. Ent., vol. 7, p. 133, 1914 (desc. fundatrix).

Records. — Chenopodvum album, C. murale; San Jose, Walnut Creek (David-
son).

94 MISCELLANEOUS STUDIES

This has been reported twice from pigweed or goosefoot in the
San Francisco Bay region, where Davidson states that it is very
common. The sexes occur in October. Davidson believes that there
is an alternate host, but as to what it might be, he is uncertain. The
author has never collected specimens, but has had access to material
taken by R. W. Doane on Chenopodium in Utah in August, 1916.

111. Aphis avenae Fabr.

Figures 191, 201, 202

Fabricius, Ent. Syst, p. 736, 1775 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 254, 1903 . Nectarophora (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. Siphocoryne (list).
Essig, Pom. Jour. Ent., vol. 3, p. 465, 1911. A. padi Linn. (list).
Essig, Pom. Jour. Ent., vol. 4, p. 790, 1912. A. maidis Fitch (desc).
Smith, Mon. Bull. Cal. Comm. Hort., vol. 3, p. 116, 1914 (list).
Davidson, Mon. Bull. Cal. Comm. Hort., vol. 6, p. 65, 1917 (note).

Records. — Graminaceae (various spp.) ; California, December to May (David-
son, Essig, Morrison, author) : PJialaris arundinacea; Stanford University, May
to July, 1915: Dracaena draco; Stanford University, June, 1915: Musa sapientum;
San Diego, March, 1916: Typha latifolia; (Davidson).

This is an abundant species through the state, occurring during
the late winter and spring on grasses and grains, migrating to other
hosts as these become ripened and dried.

The life history of this species, according to Davis (U. S. Dept.
Agr., Bull. Ill, April, 1914), is somewhat as follows:

The spring colonies on grains and grasses originate from viviparous females
which passed the winter on the grains and grasses, or from spring migrants from
the apples or related fruits; i.e., the progeny of the aphids hatching from eggs
laid the previous fall on such trees. As the weather becomes cooler they seek the
lower parts or the roots of wheat and other plants of the grass family, and
here pass the winter as viviparous females; or the winged fall migrants from the
grain may seek such trees as the apple, where the true sexes are produced.

Undoubtedly the most common method of wintering over in Cali-
fornia is on the roots and lower parts of the grains and grasses. This
species has never been collected on apples or other related trees in
this state, nor have the eggs ever been observed. During the early
spring it is found abundantly on the grains and small grasses, in
January and February in the southern part of the 'state, and during
April and May in the central part. As the grains ripen and the
stalks and leaves become hardened, it seems that the aphids migrate
to other varieties of grass which remain soft and green later, as
canary grass and reed grass and corn, or even to such hosts as the

A SYNOPSIS OF THE APHIDIDAE 95

dragon tree, cat-tail rush, and the banana. But the winter is spent
as viviparous females on the grains and grasses.

This species has been confused many times with other species
infesting grains, such as Macrosiphum granarium (Kirby) and Tox-
optera graminum (Rond.). As the latter does not occur in this state
it cannot be confused here with Aphis avenue Fabr. Clarke listed this
as Nectarophora avenue Fabr., so it appears that he might have had
Macrosiphum granarium (Kirby) in mind, as it is highly improbable
that he could have confused Aphis av&nae Fabr. with a species of
Macrosiphum (Nectarophora). The cornicles of avenae Fabr., the
absence of antennal tubercles, and the irregular venation make it
quite easily distinguishable. The cornicles are quite short, as com-
pared with a species of Macrosiphum, and distinct antennal tubercles
are entirely lacking. The third discoidal vein of the forewing is
typically twice-branched, but the second is close to the apex of the
wing, and sometimes is entirely lacking. The only other species of
Aphis in this state with this character is Aphis salicicola Thomas,
found on willows. These two are readily distinguished from each
other by the comparative lengths of the cornicles, which are consider-
ably longer in salicicola Thomas than in avenae Fabr.

112. Aphis brassicae Linnaeus

Figures 203, 204

Linnaeus, Syst. Nat., vol. 2, p. 734, 1735 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 250, 1903 (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911 (list).
Essig, Pom. Jour. Ent., vol. 3, p. 523, 1911 (desc.).

Record. — Cruciferae (various spp.) ; throughout California.

During the late winter and spring cruciferous plants are often
heavily infested with this species. Of the cultivated plants cabbages
and radishes seem to be most heavily infested ; while the wild mustard
and radish often have the entire flower clusters covered with these
aphids. Oftentimes in the colonies of this species are also found
Aphis pseudobrassicae Davis, Rhopalosiphum lactucae (Kalt.), and
R. persicae (Sulz.). In southern California the colonies are always
attacked by the braconid fly, Diaretus rapae Curtiss, and a large per-
centage of the individuals destroyed. As summer comes on these para-
sites and such predators as syrphids and ladybirds usually get the
best of the aphids, which disappear to a large extent until fall.

96 MISCELLANEOUS STUDIES

113. Aphis calendulicola Monell

Monell, U. S. Geol. Geog. Surv., Bull. 5, p. 23, 1879 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 250, 1903 (list).

Eecord. — Calendula officinale; Berkeley (Clarke).

This species has not been recognized since Clarke 's report of it on
marigold. It is possible that he had Aphis senecio Swain, which is
very common on marigolds throughout the state.

134. Aphis cardui Linn.

Figures 208, 209

Linnaeus, Syst. Nat., vol. 2, p. 735, 1735 (orig. desc.).

Games, Mon. Bull. Gal. Comm. Hort., vol. 1, p. 399, 1912. Aphis pruni

(list).

Davidson, Jour. Econ. Ent., vol. 5, p. 407, 1912 (list).
Patch, Maine Agr. Exp. Sta., Bull. 233, p. 263, 1914 (desc.).

Records. — Cirsium sp. ; San Jose (Davidson); Berkeley, June, 1915: Prunu-s
domestica; Orangevale, Sacramento County (Carnes) ; Walnut Creek (Davidson) ;
Berkeley, March, 1916 (Essig).

According to Patch this thistle aphid is the same as the one infest-
ing plums and formerly known as A. pruni Koch. Both are abundant
in the San Francisco Bay region, pruni being found in the fall and
spring on plum, cardui during the summer on thistle. The author
has attempted no transfer tests, so accepts Patch's statement as
authority for the synonymy. It is certain that structurally these are
strictty identical.

115. Aphis cari Essig

Essig, Univ. Calif. Publ. Entom., vol. 1, pp. 317-321, 1917 (orig. desc.).

Eecord. — Carum kelloggii; Eutherford, Napa County (Essig) ; Angelica
tomentosa; Berkeley (Essig).

Essig recently described this from specimens taken on wild anise
in Rutherford. The author has seen cotype specimens, but has never
collected the species.

116. Aphis ceanothi Clarke

Figures 210, 211

Clarke, Can. Ent., vol. 35, p. 250, 1903 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910 (list).
Essig, Pom. Jour. Ent., vol. 3, p. 525, 1911. Aphis ceanothi-hirsuti n. sp.
(dese.).

A SYNOPSIS OF THE APHIDIDAE 97

Records. — Ceanothus integcrrimus; Coif ax, Placer County (Clarke) ; Witch
Creek, San Diego County, June, 1916: C. cuneatus; Stanford University (David-
son), November, 1910 (Morrison), October, 1915 (R. A. Vickerey) : C. thysiflorus;
Bear Creek Gulch, Santa Clara County, April, 1911 (Morrison) : C. hirsuti; Santa
Paula (Essig).

This is a widely distributed species, having been found on Ceano-
thus as far north as Placer County, and as far south as San Diego
County. It is seldom abundant, however. The species that Essig
described as A. ceanothi -hirsuti n.sp. is undoubtedly the same as
Clarke described.

117. Aphis cerasifoliae Fitch

Figtures 288 to 292

Fitch, Eept. Ins. N. Y., vol. 1, p. 131, 1855 (orig. desc.).
Patch, Maine Agr. Exp. Sta., Bull. 233, p. 260, 1914 (desc.).

Record. — Prunus emarginata; Wynola, San Diego County, June, 1916.

This aphid was found abundantly curling the terminal leaves of
wild cherry near Wynola (3700 feet altitude), San Diego County, in
June, 1916. Alate and apterous viviparous females as well as nymphs
were abundant in the curled leaves. The apterae and nymphs were
slightly pulverulent. This species corresponds very closely to Aphis
cerasifoliae Fitch as described by Patch (op. cit.), although there are
some minor differences. Following is a copy of Patch 's description of
the Maine specimens of this species :

. This well defined species is common on both the native choke cherry, Prunus
virginiana, and the western P. demissa Walp. introduced in a nursery row on our
campus.

Apterous female. — Head, pale green or water whitish, beak short, extending
to second coxae, eyes, antennae with I, II and III concolorous with head, distal
half darker to black, III with no sensoria, proportions as shown in figure; pro-
thorax pale green, lateral tubercles present; thorax green with dark green mid-
dorsal line, femora and tibiae pale and tarsi black; abdomen pulverulent, pale
green with dark green median line and dark green transverse lines between seg-
ments, lateral tubercles present, cornicles pale with dusky tips, slender, slightly
tapering, and approximately twice the tarsus in length, cauda white with dark tip;
conical, being broad at base and abruptly tapering.

Nymphs and pupae are also pulverulent and have dark green middorsal and
transverse intersegmental line, though these are not always well defined in the
pupa which has two lateral dark green lines on thorax.

Alate female. Head black, beak short, not reaching to second coxae, eyea
black, antennae dark, III with from about 12 to 18 large sensoria about the size
of the terminal one on V, IV with from none to several sensoria like those on III,
proportions of joints as shown in the figure; prothorax green with black trans-
verse band, lateral tubercles present; thorax black, wings iridescent with slender
brown veins and large dusky stigma with pointed tip; commonly though not

98 MISCELLANEOUS STUDIES

always with second branch very short, abdomen glabrous, rather bright though not
vivid green, median line dark green, sutural lines dark green ending in marginal
green dots, cornicles dark, cauda green.

Aphis cerasifoliae is gregarious on the ventral surface of the terminal leaves
badly curling and deforming them. A copious amount of honeydew is present,
and ants are usually found attending a colony of this species.

The specimens from Wynola agree very well with this description,
although as stated above, there are a few minor points of difference.
However, as Dr. Patch writes: "It seems too close to cerasifoliae to
give it a distinct name," and "if the appearance in life answers my
description of cerasifoliae I should be inclined to call it that. It hap-
pens to be a species as characteristic alive as dead." Following are
the notes the author took of its appearance alive, before he suspected
its identity: "Alates, apterae and nymphs abundant on terminal
leaves curling them badly. Large amount of honeydew and many
ants in attendance. Apterae and nymphs pulverulent." These notes
agree exactly with Patch 's notes, cited above.

Following is a brief description of specimens taken at Wynola on
July 8 :

Apterous viviparous female. — Prevailing color pale apple green,
pulverulent. Head luteous. Thorax and abdomen pale green with
middorsal longitudinal stripe darker green. Antennae with the three
basal joints luteous, the three apical joints shading into black. Pri-
mary sensoria on V and VI, accessory sensoria on VI, no secondary
sensoria. Ill and spur are subequal, or III slightly the longer. IV
and V subequal and a little more than one-half as long as III. In
some cases IV is slightly longer than V. VI is about one-fourth as
long as its spur, longer than I, which in turn is longer than II. The
antennae are longer than the body. Cornicles long, slightly tapering,
pale with tip dusky, about equal in length to the fifth antennal seg-
ment and about twice the length of the hind tarsus. Cauda long,
conical, and about two-thirds the length of the cornicles, pale with tip
dusky. Lateral tubercles are present on the first and seventh abdom-
inal segments and on one other of the abdominal segments, in some
cases on the second, in others! on the third, and in others on the fourth.

Measurements (of specimens mounted in Canadian balsam) : Body
length, 1.5 to 1.53 mm.; body width (abdomen), 0.247 mm.; antennae
total, 1.445 to 1.734 mm. (av. 1.6082 mm.) ; I, 0.085 to 0.117 mm.
(av. 0.0987 mm.) ; II, 0.068 mm.; Ill, 0.408 to 0.467 mm. (av. 0.4335
mm.) ; IV, 0.238 to 0.306 mm. (av. 0.272 mm.) ; V, 0.221 to 0.233 mm.
(av. 0.224 mm.) ; VI, 0.1105 to 0.119 mm. (av. 0.1169 mm.) ; spur,

A SYNOPSIS OF THE APHIDIDAE 99

0.408 to 0.45 mm. (av. 0.4186 mm.) ; cornicles, 0.221 to 0.255 mm.
(av. 0.2401 mm.); cauda, 0.15 mm.; hind tarsi, 0.12 to 0.135 mm.
(av. 0.1275 mm.).

Alate viviparous female. — Prevailing color pale to apple green.
Head, antennae, thorax, marginal spots on abdomen, cornicles, tip of
cauda, femora, and tarsi all black. Antennae (fig. 289, 290) with the
usual primary sensoria on V and VI and the usual accessory sensoria
on VI. IV without sensoria and III with from 6 to 11 fairly large
circular secondary sensoria, the usual number being 8 (fig. 290). In
this character it differs most markedly from the Main specimens,
which have from 12 to 18 sensoria on III and from none to several
on IV. The antennae are slightly shorter than the body although
practically of the same length. Ill is the longest segment, closely
followed by the spur, then by IV, V, VI, I and II. Ill and the spur
are subequal, or either one or the other may be slightly the longer.
In Patch's drawing V is a little longer than IV. In the California
specimen IV is always slightly the longer of the two. In all the
California specimens the antennal segments are all a little shorter
than in the Maine material. Lateral tubercles are present on the pro-
thorax; they are always present on the seventh abdominal segment,
and may be present on any of the first few segments of the abdomen
as well. In one case they were observed on the second and seventh
segments, in another on the second, third, and seventh, in still another
on the fourth, fifth, and seventh, and in a fourth case on the first,
second, third, fourth, and seventh segments (fig. 292). The wings
and venation are normal, with the second branch of the cubitus arising
nearer to the tip of the wing than to the base of the first branch (fig.
291). However, it is not quite so close to the wing tip as in the Maine
specimens. The cornicles (fig. 292) are long and cylindrical. They
are equal to or slightly shorter than V, and from one and one-half to
two times as long as the hind tarsi. The cauda (fig. 292) is more or
less ensiform, about one-half as long as the cornicles, reaching to the
tip of the cornicles, and subequal to or slightly shorter than the hind
tarsi.

Measurements (of specimens mounted in Canadian balsam) : Body
.length, 1.53 to 1.65 mm. (av. 1.585 mm.) ; width of thorax 0.697 to
0.765 mm. (av. 0.731 mm.), antennae total, 1.568 mm.; I, 0.068 to
0.085 mm. (av. 0.0765 mm.) ; II, 0.051 mm.; Ill, 0.331 to 0.408 mm.
(av. 0.3644 mm.) ; IV, 0.238 to 0.289 mm. (av. 0.2817 mm.) ; V, 0.221
to 0.247 mm. (av. 0.2295 mm.); VI base, 0.085 to 0.111 mm. (av.

100 MISCELLANEOUS STUDIES

0.1015 mm.) ; VI spur, 0.391 mm.; cornicles, 0.204 to 0.246 mm. (av.
0.2179 mm.) ; cauda, 0.103 to 0.119 mm. (av. 0.1084 mm.) ; hind tarsi,
0.136 mm.

118. Aphis cooki Essig

Figures 212 to 214

Essig, Pom. Jour. Ent., vol. 2, p. 323, 1910. Aphis gossypii Glover (desc.).
Essig, Pom. Jour. Ent., vol. 3, p. 587, 1911 (orig. desc.).

Record. — Citrus sp., Pomona (Essig).

In 1909, C. H. Vary, county horticultural inspector in Pomona,
found a few orange trees heavily infested with this aphid. Prompt
control measures were taken and since then it has never again been
observed. Essig first thought it to be Aphis gossypii Glover and de-
scribed it under that name. Later, however, he found it to be an
undescribed species, so named it Aphis c&oki n.sp. after Dr. A. J. Cook.

119. Aphis cornifoliae Fitch

Fitch, Cat. Homop. N. Y., p. 65, 1851 (orig. desc.).

Records. — Cornus pubescens, Sanicula mensiesii; San Francisco Bay region
(Davidson).

A species comparing very favorably with this has been taken by
Davidson a number of times in the San Francisco Bay region. The
fall and winter is spent on dogwood, the summer on gambleweed.
Davidson writes as follows :

This aphid [from Sanicula] certainly appears to be very close to what I have
called (after Gillette) cornifoliae. Moreover, I have noticed that the two plants,
dogwood and Sanicula, frequently grow near each other and that there appeared
to be a migration of alates from the former just about the time there was a
migration of the alates to the latter.

This migration took place the latter part of April in 1916.

120. Aphis crataegifolii Fitch

Fitch, Cat. Homop. N. Y., p. 66, 1851 (orig. desc.).
Sanborn, Kan. TJniv. Sei. Bull. 3, p. 53, 1904 (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910 (list).

Record. — Crataegus oxycantha; San Jose, Palo Alto (Davidson).

This has been reported more or less abundant on hawthorne in the
San Francisco Bay region. According to A. C. Baker this is a good
and distinct species and not a synonym of Aphis pomi De Geer, as
formerly believed.

X

A SYNOPSIS OF THE APHIDIDAE 101

121. Aphis euonomi Fabr.

Figures 182, 187, 190, 205 to 207, 236, 237

Fabricius, Syst. Ent., p. 736, 1794 (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909. A. rumicis Linn, (list, in

part?).
Essig, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 446, 1915. A. rumicis Linn.

(list).

Becords — Althaea rosea, Berkeley, June, 1915; Hisbiscus moscheutos, Berkeley,
July, 1915 : Maytenus boaria, Berkeley, July, 1915 ; Mesembryanthemum equilat-
erale, Stanford University, June, 1915; Silybum marianum, Stanford University,
July, 1915: Urtica holoserica, Menlo Park, San Mateo County, January, 1915:
Calendula officinale, Orange, February, 1917: Antliemis sp., Pasadena, April, 1917:
Papaver sp., El Cajon, San Diego County, May, 1916 (Aphis papaveris Fabr.?) :
Vicia fdba, Stanford University (Davidson), Oxnard (Essig, 1915), Montebello,
Los Angeles County, December, 1916, Eiverside, January to May 1917 (Aphis
fabae Scop.?): Eumex spp., Palo Alto, January, 1912 (Davidson), Stanford
University, March, 1912 (Morrison), March, 1915, Ventura County, May, 1917:
Pliaseolus spp., Ventura County, May, 1917 (Aphis rumicis Linn.?).

There has been a great deal of confusion regarding the identity
of this species of aphid, and as yet its synonomy is not worked out
satisfactorily. The following is offered only provisionally by the
author. The common black aphid has usually been considered as
Aphis rumicis Linn., Aphis euonomi Fabr. being taken as a synonym,
but according to Gillette, Linnaeus' description calls for an aphid
' ' brass}1- brown in color, and not black according to the popular opin-
ion; and its food plant should be species of Eumex." He considers
the common black species to be Aphis euonomi Fabr., as does Mord-
wilko in the European form. The author follows these two aphidol-
ogists in placing Aphis rumicis Linn, of American authors (and later
European authors) as a synonym of Aphis euonomi Fabr. He (i.e.,
Gillette) writes, "whether or not it is synonymous with rumicis we
are not certain, but we very much doubt this being the case." As
long ago as 1894, Osborn and Sirrine (Iowa Agr. Sta., Bull. 26, p.
904, 1894) proved that the species which wintered in Iowa on
Euonymus migrated to Eumex and other plants in the summer. In
California the author has been unable to find it at any time upon
Euonymus, although this is a very common ornamental plant, especi-
ally in the vicinity of Riverside. This may be due, however, to the
mild winter climate of southern California, which permits plant lice to
live throughout the winter, thus not necessitating the laying of eggs.
Concerning the identity of the California species the author believes
the form described briefly below to be Aphis euonomi Fabr. The

102 MISCELLANEOUS STUDIES

one following is probably the same species, and is the one described
as Aphis papaveris by Fabricius. The species from Vicia faba is
probably the species described as Aphis fabae Scop., which may be
synonymous with Aphis eiwnomi Fabr., but again may not be. The
author tried a few transfer tests this spring (1917) with the form
from Vicia, attempting to colonize it on Hedera helix and on Rumex
spp., with negative results. Of course, this does not prove that it will
not colonize on these plants, although the author has come to the conclu-
sion that the Hedera species is entirely different, being Aphis hederae
Kalt. Dr. Patch16 in her interesting paper on aphid ecology makes
the following statement regarding migration tests, which, it seems to
the author, it is well to remember when making such tests :

If an investigator fails in one hundred attempts to colonize thistle with
migrants from plum, that will not be a safe reason for him to conclude that he is
not working with Aphis cardui, or that this thistle aphid has nothing to do with
the leaf deformations of the plum in the spring. It has been my experience that
negative data with aphids under such conditions are just no data at all. If the
structural characters are such as warrant the migration test in the first place, they
warrant a patient continuation even in the face of repeated failures.

On the other hand (and this is a most encouraging and stimulating circum-
stance in connection with aphid migration tests), a single success goes a long way
to prove the case. Barring complications, a single success is enough, and repe-
titions and verifications are needed only as safeguards in that respect.

The third description is from specimens taken on Rumex spp. and
although slightly different from the one considered as Aphis eiwnomi
Fabr., it may be the same, and it may be Aphis rumicis Linn., but of
this the author is doubtful.

In the bean fields of Ventura County, this black bean aphis is very
abundant, and often does considerable damage. In May, 1917, the
bean plants were just beginning to appear, and as yet were not
infested with the aphis. However, the native dock was quite heavily
infested. It seems that the aphis lives over the winter on dock and
perhaps on other native plants, migrating in the early summer to the
beans. Here it lives throughout the summer, returning to dock when
the beans have been harvested and the plants plowed under. Horti-
cultural Commissioner A. A. Brock, of Ventura County, places great
hope in the efficiency of Hippodamia convergens Guerin as a con-
trolling factor. In the spring of 1917 he collected a vast number of
these ladybird beetles in Sespe Canyon and turned them loose in the
bean fields just as the aphids were beginning to appear. At the
present time the results are unknown.

!« Patch, Edith M., Concerning problems in Aphid ecology, Jour. Econ. Ent.,
vol. 9, pp. 44-51, 1917.

A SYNOPSIS OF THE APHIDIDAE 103

The following brief description was made from specimens col-
lected from the first six host plants listed above, and is the one con-
sidered as Aphis euonomi Fabr.

Alate viviparous female. — Color apparently black, but on close
examination it seems that the ground color is a very dark brown,
covered with a blackish tinge, with the following parts decidedly
black: head, antennae, thoracic lobes, marginal spots and transverse
bands on the abdomen, cornicles, tarsi, coxae, tips of tibiae, and apical
one-half to two-thirds of the middle and hind femora.' The tibiae and
fore femora are pale, appearing whitish in life. The antennae are
shorter than the body, III being the longest segment, followed closely
by VI spur. In one case VI spur was slightly longer than III and
in another equal to III. In all other specimens III was the longer
segment. IV and V are subequal, V usually being slightly the
shorter. There are from eleven to twenty-one secondary sensoria on
III, of irregular size. These are scattered along the whole length of
the segment, the distal five or six being in a more or less even line.
The usual number is about twelve to fourteen. The number of
secondary sensoria on IV range from none to seven, the modal number
being two. In one specimen only were sensoria absent from IV; in
another, one antenna had seven, the other having two, while in a
third, one antenna had five, the other six. When there are more than
two or three sensoria, they are all quite small, and can be clearly
distinguished only by the higher power of a microscope. Two is the
usual number, being located about the middle of the segment. V is
usually without secondary sensoria, the primary sensorium being
always present, however. In one specimen the antennae had one or
two very small secondary sensoria on V, and in another specimen one
antenna had one small sensorium, the other none. The usual primary
and accessory sensoria are present on VI base. Lateral abdominal
tubercles are always present on the seventh segment, usually on the
first, and often on the second, third, fourth, or fifth. There are
always at least three pair of these tubercles, and oftentimes more.
One specimen had tubercles on the first, second, third, fourth, and
seventh segments. The cornicles are black, imbricated, and taper
noticeably from base to apex. They are quite constant in length, the
variation being not more than 0.05 mm. in all the specimens examined.
They are about half as long again as the hind tarsi. The cauda is
concolorous with the abdomen, short and conical or ensiform, and
subequal in length to the hind tarsi. The wings are normal, with the
typical Aphis venation.

104 MISCELLANEOUS STUDIES

Measurements: Body length, 1.53 to 1.989 mm. (av. 1.74 mm.);
width of thorax, 0.68 to 0.918 mm. (av. 0.765 mm.) ; antennae total,
1.122 to 1.36 mm. (av. 1.272 mm.) ; III, 0.289 to 0.425 m.m (av. 0.3648
mm.) ; IV, 0.1955 to 0.272 mm. (av. 0.2266 mm.) ; V, 0.187 to 0.221
mm. (av. 0.1885 mm.) ; VI, base 0.102 to 0.136 mm. (av. 0.1119 mm.) ;
VI, spur 0.289 to 0.357 mm. (av. 0.3145 mm.) ; cornicles, 0.1785 to
0.221 mm. (av. 0.2118 mm.) ; cauda, 0.136 to 0.162 mm. (av. 0.14875
mm.) ; hind tarsus, 0.136 to 0.152 mm. (av. 0.1372 mm.).

Specimens taken by the author in May, 1916, on Papaver sp. (cul-
tivated poppy) near El Cajon, San Diego County, seem to him to be
Aphis papaveris Fabr. (Genera Insectorum, p. 303, 1717), and prob-
ably are the same as the above species, although they may be different.
There are from thirteen to fifteen irregular secondary sensoria on III
as above, but IV and V are without secondary sensoria, with one
exception, in which there was one small sensorium near the middle of
IV. The cauda is equal to the hind tarsi, the cornicles being longer,
and about the same comparative length as above. The third antennal
segment appears to be longer in comparison than above in some speci-
mens. Lateral abdominal tubercles are present on the first, third, and
seventh abdominal segments.

Measurements: Body length, 1.486 to 1.908 mm. (av. 1.711 mm.) ;
width of thorax, 0.595 to 0.765 mm. (av. 0.68 mm.) ; antennae total,
1.224 to 1.343 mm. (av. 1.2878 mm.) ; III, 0.323 to 0.374 mm. (av.
0.3536 mm.) ; IV, 0.2125 to 0.22 mm. (av. 0.2193 mm.) ; V, 0.187 to
0.204 mm. (av. 0.2024 mm.) ; VI, base 0.102 to 0.119 mm. (av. 0.1054
mm.) ; VI, spur 0.255 to 0.34 mm. (av. 0.2992 mm.) ; cornicles, 0.187
to 0.221 mm. (av. 0.204 mm.) ; cauda, 0.136 to 0.153 mm. (av. 0.142
mm.) ; hind tarsus, 0.119 mm.

Specimens taken by the author near Montebello, Los Angeles
County, in December, 1916, and in Riverside from January to May,
1917, on Vicia faba seem to be somewhat different from the fore-
going, yet are very nearly identical. Gillette considers that they
might possibly be Aphis fdbae Scop., which may or may not be
synonymous with Aphis euonomi Fabr. Superficially, the coloring
seems to be the same, although on close observation it appears to be
a very dark green in ground color, covered with a blackish tinge. The
legs are colored as above, however.

Specimens from Rumex appear to have considerably more brown
in the ground color than the preceding varieties. Secondary sensoria
are located as follows: III, 14 to 24 (av. 18) ; IV, 4 to 7 (av. 5) ; V,

A SYNOPSIS OF THE APHIDIDAE 105

I to 4 (av. 3). Lateral abdominal tubercles could be found only on
the first and seventh segments.

Alate viviparous female. — Measurements: Body length, 1.768 to
2.142 mm. (av. 1.942 mm.) ; width of thorax, 0.782 to 1.054 mm.
(av. 0.918 mm.) ; antennae total, 1.445 to 1.581 mm. (av. 1.496 mm.) ;
III, 0.357 to 0.408 mm. (av. 0.394 mm.); IV, 0.255 to 0.323 mm.
(av. 0.286 mm.) ; V, 0.204 to 0.255 mm. (av. 0.233 mm.) ; VI, base
0.136 to 0.153 mm. (av. 0.139 mm.) ; VI, spur 0.289 to 0.323 mm.
(av. 0.306 mm.) ; cornicles, 0.187 to 0.255 mm. (av. 0.219 mm.) ; cauda,
0.136 to 0.17 mm. (av. 0.153 mm.) ; hind tarsus, 0.119 to 0.153 mm.
(av. 0.147mm.).

Apterous viviparous female. — Measurements : Body length, 2.278 to
2.448 mm. (av. 2.3403 mm.) ; antennae total, 1.309 to 1.598 mm.
(av. 1.4382 mm.) ; III, 0.306 to 0.408 mm. (av. 0.3502 mm.) ; IV,
0.221 to 0.306 mm. (av. 0.2618 mm.) ; V, 0.206 to 0.255 mm. (av. 0.238
mm.) ; VI, base 0.119 to 0.17 mm. (av. 0.1394 mm.) ; VI, spur 0.289 to
0.34 mm. (av. 0.306 mm.) ; cauda, 0.17 to 0.204 mm. (av. 0.187 mm.) ;
cornicle, 0.255 to 0.323 mm. (av. 0.289 mm.) ; hind tarsus, 0.153 to
0.17 mm. (av. 0.167 mm.).

122. Aphis frigidae Oestlund

Oestlund, Geol. Nat. Hist. Snrv. Minn., vol. 14, p. 46, 1886 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. 132, 1913 (desc. stem mother).

Eecords. — Artemisia californica; Walnut Creek, Contra Costa County (David-
son).

In company with Macrosiphum artemisiae (Fonsc.) this species is
found on sagebrush in the San Francisco Bay region. Wilson reports
it from Oregon, so probably it is distributed along the coast from the
bay north. In the course of observations in southern California
during a period of two years the author has been unable to find any
aphids infesting sagebrush.

123. Aphis gossypii Glover

Figures 192, 193, 215

Glover, Pat. Off. Eec., p. 62, 1854 (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 250, 1903 (list).

Essig, Pom. Jour. Ent., vol. 1, p. 47, 1909. Aphis citri Ashmead (desc.).

Essig, Pom. Jour. Ent., vol. 3, p. 590, 1911 (desc.).

Cook, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 65, 1912 (list).

Games, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 398, 1912 (list).

Weldon, Mon. Bull. Cal. Comm. Hort., vol. 2, p. 597, 1913 (list).

Davidson, Mon. Bull. Cal. Comm. Hort., vol. 6, p. 65, 1917 (note).

106 MISCELLANEOUS STUDIES

Records. — Cucumis spp. ; Newcastle, Placer County, Watsonville, Santa Cruz
County (Clarke); Imperial County (Weldon) ; San Diego County, June, 1916:
Cucurbita spp.; Alpine, San Diego County, June, 1916: Citrus spp.; Santa Paula,
Claremont (Essig), Acampo, San Joaquin County (Games), San Diego, March,
1916 (E. E. McLean); Whittier, May, 1917: Heracleum lanatum; Berkeley,
March, 1915 (Essig): Begonia; Stanford University, February, 1912 (Morrison),
Biverside, January, 1917; Punica granatum, Stanford University, April, 1911
(Davidson): Helianthus; Santa Ysabel, San Diego County, May, 1916: Pcrsea
gratissima; Avondale, San Diego County, August, 1916; Chrysanthemum;
Ontario, January, 1917; Esclischoltzia calif omica ; Ontario, January, 1917:
Anthemis spp.; Pasadena, April, 1917 (E. E. Campbell): Pyrus spp.; Santa
Cruz County (Volck), Nevada County (Norton).

The melon or cotton aphis is distributed throughout the state and
is found on a large number of host plants. On melons it is often a
considerable pest, particularly in the Imperial Valley. In the apple
sections of Santa Cruz and Nevada counties it often becomes abundant
enough upon the young trees to cause considerable damage, according
to County Horticultural Commissioners Volck and Norton. In San
Diego County the author found an infestation on young avocado trees
which was very severe. Oftentimes it becomes quite abundant in
nurseries and greenhouses.

124. Aphis hederae Kalt.

Kaltenbach, Monog. d. Pflanzenlause, p. 89, 1843 (orig. dese.).

Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909. A. rumicis Linn, (list in

part).

Essig, Pom. Jour. Ent., vol. 2, p. 335, 1910 (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910. A. rumicis Linn. (list).

Records. — Hedera helix; Stanford University (Davidson), March, 1912 (Mor-
rison) ; Claremont, Los Angeles County (Essig) ; San Jose, May, 1911 (Davidson,
Morrison); Oakland, November, 1916 (Davidson); Berkeley, April, 1915; Lemon
Grove, San Diego County, March, 1916; Eiverside, October, 1916: Chcnopodium
sp., Walnut Creek, Contra Costa County, May, 1915 (Davidson).

Throughout the San Francisco Bay region and southern Califor-
nia a small dark brown to black aphid is often found in colonies on
the tender shoots of English ivy. Essig described it as Aphis hederae
Kalt., but later it was believed to be Aphis rumicis Linn. (A. euonomi
Fabr.). However, a careful study of a large series of specimens of
this aphid from ivy and of A. euonomi Fabr. from a number of dif-
ferent host plants has convinced the author that they are distinct.
Gillette is of the same opinion. Consequently the species from ivy
in California is Aphis hederae Kalt. In the author's collection there
is a specimen from Chenopodium sp. taken by Davidson that appears
to be the same species. The most noticeable difference between this

A SYNOPSIS OF THE APH1DIDAE 107

and Aphis euonomi Fabr. is in the length of the cornicles, which are
very much longer in this species. Measurements of specimens of the
alates from Oakland, Walnut Creek, San Jose, and Riverside are
herewith given:

Measurements: Body length, 1.411 to 1.768 mm. (av. 1.621 mm.) ;
width of thorax, 0.714 to 0.782 mm. (av. 0.748 mm.) ; antennae total,
1.411 to 1.549 mm. (av. 1.499 mm.) ; III, 0.323 to 0.391 mm. (av. 0.365
mm.) ; IV, 0.272 to 0.323 mm. (av. 0.2914 mm.) ; V, 0.221 to 0.272 mm.
(av. 0.2518 mm.) ; VI, base 0.119 to 0.136 mm. (av. 0.311 mm.) ; VI,
spur 0.306 to 0.34 mm. (av. 0.323 mm.) ; cauda, 0.136 mm.; cornicle,
0.306 to 0.34 mm. (av. 0.3252 mm.) ; third tarsus, 0.119 to 0.136 mm.
(av. 0.1237 mm.).

It will be seen that the cornicles are considerably more than twice
as long as the hind tarsi, in some cases practically three times, while
in A. euonomi Fabr., they are scarcely twice as long as the hind tarsi.
In A. euonomi Fabr. the hind tarsi are longer than the base of VI,
while the cornicles are shorter than VI spur. In A. hedera-e Kalt. VI
spur and the cornicles are subequal or on the average the cornicles are
very slightly longer, while VI base and the hind tarsi are also sub-
equal, the tarsi being shorter on the average. The secondary sensoria
in A. hedcrae Kalt. are small, irregular in size, and are scattered more
or less irregularly along III but in a fairly even row along IV and V.
They appear very much the same as in A. euonomi Fabr. There are
from thirteen to twenty on III, seventeen being the average; from
five to nine on IV, seven and eight being the usual number; and
usually one on V, although in a few cases there appear to be none.

125. Aphis heraclei Co wen

Cowen, Hemip. Colo., p. 120, 1895 (orig. dese.).

Essig, Univ. Calif. Publ. Entom., vol. 1, p. 339, 1917 (list).

Record. — Heracleum montezzamum ; Berkeley (Essig).

Recently Essig reported having taken this species on Heracleum
in Berkeley. The author has specimens from Essig, although he has
never collected it himself. This is the only report of the species since
Cowen 's original report and description.

126. Aphis houghtonensis Troop?

Troop, Ent. News, vol. 17, p. 59, 1906 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. 132, 1914 (list).

Eecord. — Ribes sanguineum; Contra Costa County (Davidson).

108 MISCELLANEOUS STUDIES

Davidson reported a species of Aphis infesting the terminal leaves
of wild currant in the canyons of Contra Costa County. He identified
it provisionally as this species as he was uncertain. The author is
unacquainted with it.

127. Aphis lithospermi Wilson

Wilson, Trans. Am. Ent. Soc., vol. 41, p. 100, 1915 (orig. desc.).
Record. — Lithospermum pilosum; California (Wilson).

There is no definite record of this species in California, but it is
listed here because Wilson added it to a list of the California Aphi-
didae submitted to him by the author.

128. Aphis maidis Fitch

Figures 216 to 218

Fitch, Insects N. Y., vol. 1, p. 318, 1855 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 251, 1903 (list).
Davidson, Jour. Econ. Ent., vol. 5, p. 408, 1912 (list).

Records. — Corn; Watsonville, Berkeley (Clarke); San Jose (Davidson); Lake-
side, San Diego County, April, 1916; Chula Vista, San Diego County, August,
1916: sorghum; Julian, San Diego County, August, 1916 (H. M. Armitage) ;
Corona, Eiverside County, September, 1916.

Only occasionally is this corn aphis found in California, where it
infests the ears and tassels and leaves of corn and some of the sor-
ghums. Never has it been observed as injurious as is sometimes
reported from the middle western states.

129. Aphis malifoliae Fitch

Figures 248 to 250

Fitch, Trans. N. Y. State Agr. Soc., vol. 5, p. 14, 1854 (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 252, 1903. Aphis sorbi Ka.lt. (list).
Games, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 400, 1912. A. sorbi Kalt.

(list).
Weldon, Mon. Bull. Cal. Comm. Hort., vol. 3, p. 188, 1914. A. sorbi Kalt.

(list).
Baker and Turner, Jour. Agr. Ees., vol. 7, pp. 321-343, 1916 (complete

account). •

Records. — Pyrus mdlus, P. communis; Central and northern California; Orange
County, May, 1917.

A SYNOPSIS OF THE APHIDIDAE 109

This is one of the most injurious of our California species of Aphis,
being found in practically all of the apple-growing regions of the
state, and in most of them necessitating some control measures. It
has been reported on apple and pear in the following counties : Hum-
boldt, Orange, Placer, Sacramento, Santa Clara, Shasta, Tehama,
Nevada, Inyo, Santa Cruz, and Alameda. Probably it is present
wherever apples are grown, with the exception of the southern Cali-
fornia districts where it has never been observed. The apple is the
primary host, and only occasionally has it been taken on pear. In
May, 1917, Roy K. Bishop found it in Orange County, this being the
first report of it south of the Tehachapi.

The life history of this Aphis in California is as follows :
In the fall and early winter the eggs are laid in the crotches of the
twigs. These hatch in the following spring, the exact time depending
upon the weather conditions but it is usually as the buds are begin-
ning to show green, or as they are beginning to open. The author has
observed the young stem mothers on the young buds of the apple in
the latter part of March, although he has never been able to find the
eggs, either those yet unhatched or those from which the stem mothers
have already hatched. Horticultural Commissioner Weatherby of
Humboldt County writes that he has found the eggs hatching as early
as February 24. He goes on to state that the eggs of Aphis pomi
De Geer do not hatch until considerably later. Horticultural Com-
missioner Norton of Nevada County has made the following observa-
tions :

The eggs of Aphis sorbi [malifoliae] are laid on the buds, or sometimes on
the spurs close to the buds. At first they are hard to see as they are small and
light green, but later they turn to a shiny black, when they can be more readily
detected. The young aphids hatch as soon as the buds begin to swell, which time
varies with the season. I have found them sometimes as early as the first of
March and at other times as late as the middle of April.

The stem mothers feed upon the plant juices through the buds,
sometimes appearing on the outer surface of the buds and at other
times crawling down into the unfolding leaves, as is the case with
Aphis pomi De Geer. In a few weeks these are mature and begin to
deposit live young. All of this second generation are apterous females
so far as the author has been able to observe. On April 12, 1915, he
found several colonies of these aphids in the apple orchard at Stanford
University, each colony consisting of a stem mother and several young
apterous viviparous females. These females mature in a few weeks

110 MISCELLANEOUS STUDIES

and a third generation is begun. The most usual place to find the
second and third generations is in the curled terminal leaves of the
plant. These leaves are curled very similarly to those by the green
apple aphis (Aphis pomi De Geer), but they are curled a great deal
tighter. Winged females may appear in this third generation, but
it is most usual to find them in the fourth. Horticultural Commis-
sioner Volck of Santa Cruz County states that he has counted four
generations before the summer migration. During May, 1915, the
author collected many colonies of this Aphis and placed them in vials
in the laboratory. Many others he attempted to colonize on some
apple seedlings. Owing to various causes he was unable to make any
successful colonizations on the apple trees, one of the chief causes
being the destructive work of coccinellid larvae. Also during the
first few days of June he was forced to be absent from town and on
his return found that the gardener had "cleaned" the trees, for
"they were all covered with lice." Until May 25 no alate females
had been found, but on that date two appeared in the laboratory. On
May 10, 1917, alates were found in Orange County.

These alate females of the fourth (perhaps sometimes they appear
in the third) generation migrate from the apple to some unknown
host. At Stanford University in 1915 the migration began about
the first of June and continued for some two or three weeks. On
June 20 only two or three colonies, each consisting of but a very few
individuals, were found where a month before there had been literally
hundreds. The curled leaves still hung on the trees and in each
curled leaf the moulted skins of the aphid were abundant. From
Commissioner Norton of Nevada County comes the statement that he
has known the migrants "to leave the trees as early as the middle
of June, but the migration usually takes place between the first and
the fifteenth of July. "Where they go I have never been able to find
out, as I have never observed them on any other host plant. ' '

According to O. E. Bremner, Horticultural Commissioner of
Sonoma County, the migration takes place there during June. This
is the same as in Santa Clara County. In Orange County in 1917
the alate females appeared about the first of May. Migration began
almost immediately and continued for two or three weeks. By May 24
only a very few aphids remained. This is fully a month earlier than
migration takes place north of the Tehachapi. Incidentally the spring
of 1917 was exceedingly cool and the summer very late. In normal
years one would expect the aphids to leave the apple two or three
weeks earlier.

A SYNOPSIS OF THE APHIDIDAE 111

The summer host plant of this aphid is as yet unknown in Cali-
fornia. During June, 1915, the author spent many hours in search
of this host plant, but to no avail. He examined every kind of plant
within two or three hundred yards of the apple orchard at Stanford
University, but on none was he able to find any aphid that could pos-
sibly be the summer form of Aphis malifoliae Fitch. Bremner reports
having found isolated individuals on pigweed (Amaranthus retro-
flex'us) in Sonoma County, but believes this to be accidental for he has
never observed them to deposit young on this plant. Davidson writes
that he has been able to colonize them in the laboratory on the leaves
of plantain (Plantago spp.), in fact has been able to have them repro-
duce in such large numbers as to kill the plants. On May 28, 1915,
the author placed two alate females from apple leaves on each of two
specimens of Plantago hirtella under bell jars in the laboratory at
Stanford University. On returning to town on June 10 he found that
the plants were in a dying condition, owing to a lack of proper care
during his absence. However, he found many young lice present, all
of which were alive and feeding. The adult alate females had already
died. By June 16 the lice had moulted once, but then the plants were
practically dead. He left Stanford within a few days not to return,
so was unable to begin fresh experiments along this line. In his search
for the alates in the field he was particularly careful to examine
closely every plantain plant in the vicinity, but could find no trace of
this aphid on them. Davidson also reports the same lack of success.
Consequently, although the alates will deposit young on plaintain
in the laboratory it cannot very well be the natural summer host in
this state. Baker and Turner have proven that Plantago lanceolata
is the summer host in Virginia. W. H. Britain has observed a definite
migration to plaintain in Nova Scotia (Proc. Ent. Soc. Nova Scotia,
vol. 1, pp. 16-30, 1915). Incidentally he has been able to breed it
throughout the summer on apple. In Orange County, in the vicinity
of the known infestations, the author was unable to find any plaintain
whatsoever. On inquiring of Roy K. Bishop, the county horticultural
commissioner, it was learned that plaintain is very scarce in that
county, except very near to the coast, and that it is exceedingly doubt-
ful if there is any in the vicinity of the known aphid infestations.

The fall migrants begin to return to the apple some time during
the fall and deposit living males and females. From Nevada County
comes the report that the migrants return to the apple "between the
twentieth of September and the first of October. ' ' Davidson has taken
the oviparous females and the alate males on December 5 (1912) at

112 MISCELLANEOUS STUDIES

Sebastopol; Morrison has taken the sexes at Stanford University on
December 16 (1910) ; Moznette of the Oregon station has taken the
migrants as late as the middle of November at Corvallis, Oregon.
Consequently, egg laying probably occurs from the middle of October
well into December in the various parts of California. Commissioner
Norton states: "The first eggs that I have seen were observed about
the fifteenth of October. However, they continue egg laying, in
favorable years, well along into November. ' '

The injury caused by this aphid is done entirely in the spring of
the year, before the summer migration, and consists in the curling of
the terminal leaves. The colonies are found usually in the leaves
surrounding a cluster of apples, and although most of the feeding is
on the leaves themselves oftentimes they feed upon the fruit. In
such a case the fruit (according to Weldon, ''Apple Growing in Cali-
fornia," Mon. Bull. Cal. Comm. Hort., p. 86, 1915) "is injured to
such an extent that it becomes stunted and not only fails to mature,
but is distorted so badly that' the variety may not be recognizable. ' '
In Nevada County, Commissioner Norton reports: "The purple aphis
unless controlled lessens the apple crop from ten to fifteen per cent. ' '
This is a higher percentage, undoubtedly, than is common throughout
the state, but it shows how serious the pest may be.

130. Aphis marutae Oestlund

Figures 293 to 299
Oestlund, Minn. Geol. Nat. Hist. Surv., vol. 14, p. 40, 1886 (orig. desc.).

Eecords. — Silybum marianum; Grossmont, San Diego County, April, 1916:
Centaurea melitensis; El Cajon, San Diego County, May, 1916.

In April, 1916, the author observed a small aphid on milk thistle
near Grossmont, San Diego County, and later on tacalote in the
El Cajon Valley. It infested the smaller leaves, the leaf petioles, and
the base of the flowers. Large numbers of ants were in attendance,
but it was preyed upon extensively by the larvae and adults of Cocci-
nella California. A considerable number of adults of Lysiphlebus
testaceipes Cresson were reared from colonies of this aphid. Being
unknown to the author specimens were sent to J. J. Davis and E. 0.
Essig, both of whom determined the species to be Aphis marutae Oest-
lund. Inasmuch as Oestlund 's descriptions are the only ones avail-
able, a brief description is given below of specimens taken May 1,
1916, on Silybum marianum in San Diego County.

A SYNOPSIS OF THE APHIDIDAE 113

Alate viviparous female. — Prevailing color pale to olive green.
Head and prothorax dark olive green, thoracic lobes almost black.
Abdomen pale green with marginal spots and patch on dorsum dusky.
Legs pale except tarsi, apex of tibiae, and apical two-thirds of femora.
Antennae, cornicles, and cauda dusky. Beak pale at base and dusky
at tip.

Head (fig. 293) not quite as long as broad, with a prominent
tubercle at apex of front and small but distinct projections from head
on inner side of first antennal segments. Antennae about same length
as body or slightly longer or slightly shorter (figs. 294-295). Ill and
the spur are about equal or III slightly longer, never shorter than
spur. IV about one-half as long as III. V either shorter or equal to
IV. VI shorter than V and about one-third as long as spur. I and II
subequal and slightly shorter than VI. The usual primary sensoria
are present on V and VI and the accessory sensoria on VI. Ill is
tuberculate and IV is slightly so. IV has from two to six small,
circular secondary sensoria and III from eleven to fifteen irregularly
placed (fig. 294). The beak reaches considerably beyond the second
coxae, in some cases almost to the third.

The prothorax is without lateral tubercles. The wings are about
twice as long as the body with normal venation. The stigmal vein is
curved its entire length, the second branch of the cubitus arises about
midway between the tip of the wing and the base of the first branch.

The abdomen is without lateral tubercles in so far as the author
can discern. The cornicles (fig. 299) are short and taper slightly
from base to apex. They are about equal in length to the third tarsi,
are almost one-half as wide at base as long, and about one-third
as wide at apex as long. The cauda (fig. 298) is short and blunt
(conical) and about two-thirds as long as the cornicles. The anal
plate is half-moon-shaped and dusky at its distal edge.

Measurements (of specimens in Canada balsam) : Body length,
0.918 to 1.02 mm. (av. 0.9248 mm.) ; width (thorax), 0.34 to 0.442 mm.
(av. 0.4082 mm.) ; antennae total, 0.885 to 1.02 mm. (av. 0.942 mm.) ;
I, 0.034 to 0.051 mm. (av. 0.037 mm.) ; II, 0.034 to 0.051 mm. (av.
0.048 mm.) ; III, 0.225 to 0.2975 mm. (av. 0.2601 mm.) ; IV, 0.117 to
0.17 mm. (av. 0.152 mm.) ; V, 0.1105 to 0.136 mm. (av. 0.1346 mm.) ;
VI, 0.068 to 0.102 mm. (av. 0.0833 mm.) ; spur, 0.204 to 0.272 mm.
(av. 0.2295 mm.) ; cornicles 0.0850 to 0.119 mm. (av. 0.0978 mm.) ;
cauda, 0.0595 to 0.068 mm. (av. 0.0624 mm.) ; hind tarsi, 0.085 to
0.102 mm. (av. 0.0901 mm.) ; wing length, 1.921 to 1.955 mm. (av.
1.928 mm.) ; wing width, 0.661 mm.; wing expansion, 4.556 mm.

114 MISCELLANEOUS STUDIES

Apterous viviparous female. — The apterae are quite similar to the
alates except that the thorax is not dark, and that the second, third,
and basal three-fourths of the fourth antennal segments are pale.
There are no secondary sensoria (fig. 296) and no lateral tubercles
on prothorax and abdomen (fig. 297). The individuals are slightly
larger and the proportions of the antennal segments differ slightly
from the alates. The measurements of specimens mounted in Canada
balsam are as follows :

Measurements: Body length, 1.00 to 1.04 mm. (av. 1.026 mm.);
width (abdomen), 0.595 to 0.629 mm. (av. 0.6064 mm.) ; antennae
total, 0.561 to 0.697 mm. (av. 0.6151 mm.) ; III, 0.102 to 0.136 mm.
(av. 0.1218 mm.); IV, 0.0765 to 0.1105 mm. (av. 0.0906 mm.); V,
0.068 to 0.085 mm. (av. 0.0765 mm.) ; VI, 0.595 to 0.0765 mm. (av.
0.068 mm.) ; spur, 0.1615 to 0.1785 mm. (av. 0.1711 mm.) ; cornicles,
0.0765 to 0.11 mm. (av. 0.0935 mm.) ; cauda, 0.0595 mm.; hind tarsi,
0.102 mm. (Description from nine specimens of apterae). It will
be noticed that in the apterae the antennae are but about two-thirds
as long as the body, while in the alates they are almost as long as the
body. Furthermore, in the apterae the spur of the sixth antennal
segment is always longer than III while in the alates it is equal to III
at the most, and in many cases shorter.

131. Aphis medicaginis Koch

Figure 189

Koch, Die Pflanzenlause, p. 94, 1854 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 376, 1910 (list).
Essig, Pom. Jour. Ent., vol. 3, p. 527, 1911 (dese.).

Records. — Medicago hispida; Stanford University (Davidson), April, 1914
(R W. Haegele) : Astragalus leucopsis; Nordhoff, Ventura County (Essig) : Vicia
faba, lima bean, Pasadena (E. E. Campbell).

This small dark Aphis has been found occasionally in California,
particularly on alfalfa and beans. Such other plants as loco weed,
licorice, sagebrush, locust, and others are said to be hosts. The author
has never collected it himself, but has had access to specimens taken
by Essig, Haegele, and Campbell. Davidson has reared the braconid
fly, Lysipheebus testaceipes Cresson, from this aphid.

A SYNOPSIS OF THE APHIDIDAE 115

132. Aphis middletonii Thomas

Figures 219, 220
Thomas, 8th Ann. Kep. 111. St. Ent., p. 99, 1879 (orig. desc.).

Recards. — Amaranthus retro flexus ; Santa Paula, August, 1911 (Essig) : Ban-
unculus calif ornicus; Julian, San Diego County, June, 1916: Hemizonia rudis;
Stanford University, 1916 (Ferris) : Helianthus annuus; Kiverside, September,
1916.

In the fall of the year this species is rather common on the roots
of various plants in California. The individuals are small green
aphids, covered with a slight pulverulence. They are very similar
to Aphis maidis-radicis Forbes, with which they have often been con-
fused, and differ particularly in the presence of secondary sensoria
on the fourth antennal segment of the apterae. Below are a few
descriptive notes taken from specimens mounted in balsam, collected
in 1916 in Julian and Riverside, and in 1911 near Santa Paula:

Alate viviparous female. — Greenish, pruinose. Head, antennae,
thorax, marginal spots on abdomen, cornicles, cauda, apical one-half
femora, apices tibiae, tarsi, and apex of beak, black. Antennae reach
to the base of the second abdominal segment; III being the longest
segment, followed by VI spur. IV and V are subequal, VI base
slightly shorter. The usual primary and accessory sensoria are pres-
ent. Secondary sensoria occur on III and IV (fig. 220). There are
nine to twelve on III, and one to four on IV. The average numbers
are eight and two respectively. The beak reaches to the third coxae.
Prominent lateral tubercles are present on the first and seventh
abdominal segments, as well as on the prothorax. The cornicles are
short and taper slightly toward the apex. They are subequal in
length to the hind tarsi, and very slightly larger than the cauda. The
wings are normal, with the second branch of the third discoidal arising
nearer to the apex of the wing than to the base of the first branch.

Measurements: Body length, 1.65 to 1.7 mm. (av. 1.674 mm.);
width of thorax, 0.561 mm. ; antennae total, 0.816 to 0.918 mm. (av.
0.884 mm.) ; III, 0.204 to 0.255 mm. (av. 0.2338 mm.) ; IV, 0.11 to
0.119 mm. (av. 0.1169 mm.) ; V, 0.11 to 0.136 mm. (av. 0.1275 mm.) ;
VI, base 0.085 to 0.102 mm. (av. 0.0986 mm.) ; VI, spur 0.204 mm.;
cauda, 0.102 mm.; cornicles, 0.1275 to 0.136 mm. (av. 0.1332 mm.) ;
hind tarsus, 0.119 to 0.136 mm. (av. 0.1303 mm.) ; wing length, 1.904
to 2.38 mm. (av. 2.159 mm.) ; width, 0.731 to 0.85 mm. (av. 0.815
mm.) ; expansion, 4.3 to 5.1 mm. (av. 4.717 mm.).

116 MISCELLANEOUS STUDIES

Apterous viviparous female. — These are very similar to the alate
females, only slightly larger. The antennae are dusky throughout
except the base of III. They reach to the base of the first abdominal
segment. Ill is the longest segment. VI spur is next, being about
two-thirds as long. IV, V, and VI base are subequal, with V some-
what shorter than the others. The usual primary and accessory sen-
soria are present on V and VI. Ill has two or three small secondary
sensoria located in the apical one-third of the segment. IV has from
one to three in the apical one-half. The prothorax and the first and
seventh abdominal segments each have a pair of conspicuous lateral
tubercles. The cornicles are black and somewhat larger than in the
alates, being slightly longer than the hind tarsi. The cauda is a
little shorter than the hind tarsi.

Measurements: Body length, 1.632 to 1.785 mm. (av. 1.708 mm.) ;
width of thorax, 0.748 to 0.85 mm. (av. 0.799 mm.) ; antennae total,
0.867 to 0.969 mm. (av. 0.9265 mm.) ; III, 0.2465 to 0.289 mm. (av.
0.2635 mm.) ; IV, 0.102 to 0.136 mm. (av. 0.119 mm.) ; V, 0.102 to
0.119 mm. (av. 0.1105 mm.) ; VI, base 0.119 mm.; VI, spur 0.1615 to
0.187 mm. (av. 0.17 mm.) ; cornicles, 0.153 to 0.17 mm. (av. 0.1615
mm.) ; cauda 0.119 mm. ; hind tarsus, 0.136 mm.

133. Aphis mori Clarke

Clarke, Can. Ent., vol. 35, p. 251, 1903 (orig. desc.).
Record. — Morus sp., Berkeley (Clarke).

This is a rather doubtful species, described by Clarke from speci-
mens taken on mulberry in Berkeley. Since the original description
it has never again been observed.

134. Aphis neomexicana Ckll. var. pacifica Dvdn.

Figures 300, 302

Davidson, Jour. Econ. Ent., vol. 10, p. 293, 1917 (orig. desc. var.).
Records. — Eibes rubrum; Walnut Creek, Contra Costa County, and San Jose
(Davidson).

Davidson described this variety from specimens found curling the
leaves of cultivated red currant in Walnut Creek in June, 1915.
What he takes to be the same species he had already collected in San
Jose in May, 1912. The author has specimens from him, but has never
collected any himself.

117

135. Aphis nerii Fonsc.

Figures 221, 222

Boyer de Fonscolombe, Ann. Ent. Soc. France, vol. 10, p. 167, 1841 (orig.

desc.).

Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910 (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. A. lutescens Monell (list).
Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911. A. lutescens Monell (list).
Essig, Pom. Jour. Ent., vol. 3, p. 401, 1911. A. lutescens Monell (desc.).
Essig, Pom. Jour. Ent., vol. 3, p. 530, 1911 (dese.)
Branigan, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 53, 1915 (list).

Eecords. — Asclepias mexicana; Stanford University (Davidson) ; Stanford
University, October, 1910 (Morrison) ; Penryn, Placer County (Davidson) ; south-
ern California (Essig); Berkeley, July to September, 1915: Nerium oleander;
southern California (Essig) ; Sacramento (Branigan) ; Berkeley, August to Decem-
ber, 1915 j San Diego, 1916.

In the late spring, summer, and early fall milkweeds throughout
the state are often seen to be infested with a bright yellow and black
aphid. In the fall and early winter this same species is found infest-
ing oleanders. Where oleanders are present but no milkweeds this
aphid can be found from spring until winter on the oleander, as
observed during 1916 in San Diego.

Heretofore the species on oleander and milkweed have been con-
sidered as distinct, the former being called A. lutesc&ns Monell, the
latter A. nerii Fonsc. According to a note from J. J. Davis the
species on milkweed could not be A. lutescem Monell. Following are
extracts from his letters concerning this point :

I am wondering whether you have ever found winged specimens on Asclepias
that do not bear the black markings at the base of the cornicles. All the speci-
mens that I have collected and which Mr. Monell has collected in recent years have
these black markings at the base of the cornicles in the winged forms. However,
in referring to an old note from Mr. Monell, he says that it would seem hardly
possible that he could have missed these dark spots if they had been present in
the specimens from which he drew his description for Aphis lutescens, and re-
marks further that he is not sure that he has ever seen A. lutescens alive since he
first described it. I am wondering if lutescens is not really asclepiadis of Pass-
erini and whether our other common species on Asclepias and Nerium is not
nerii Fonsc.

During the summer of 1915 the author found this species on
Asclepias in the Botannical Gardens at the University of California.
During July and August it was quite abundant; in fact, it was
especially thick on the stems and undersides of the leaves and blossoms.
However, in the latter part of August it seemed to be getting less

118 MISCELLANEOUS STUDIES

and less numerous. No sign of parasites was present, and the pre-
daceous enemies were not more abundant than usual, so a search for
the cause was made. Within fifty feet of the milkweed plants several
oleanders were found and on them was noticed a large yellow species
of Aphis. This supposedly was Aphis nerii Fonsc. In the laboratory
the author could find no structural difference whatsoever between this
species and the one on Asclepias, so he continued to watch them care-
fully on the hosts. As the days passed the Asclepias became freer and
freer of the infestation, while the Nerium became more and more
heavily infested. This continued through September and into October,
by which time the Asclepias had died down and incidentally no aphids
were left. The Nerium was very heavily infested then. This was
taken as a good proof that these were the same species. Later Essig
told the author that the summer before (1914) he had made transfer
tests in the laboratory of specimens from Asclepias to Nerium and
that they thrived there and bred well. This fact and the observations
above mentioned were noted in a letter to Davis. Following is his
answer :

I have your letter relative to Aphis asclepiadis and nerii, and am interested
in your observations. In 1'914, Theobald described a species under the name of
Aphis nigrepes, which he now places as a variety of asclepiadis. He considers
nerii as distinct from asclepiadis because the latter lacks the black patches at
the base of the cornicles. Passerini's asclepiadis is entirely different from Fitch's
Aphis asclepiadis. Fitch's name has priority for, as you will notice, it was
described in 1851. This being the case, Passerini's name will have to fall and
be replaced by Aphis lutescens of Monell, which according to Mr. Monell 's data
does not bear the black patches around the base of the cornicles.

This would seem to indicate that the California species on Asclepias
is Aphis nerii Fonsc. and not A. lutescens Monell, as brought out by
Essig 's experiment and by the author's observation. Consequently
this Californian species is Aphis nerii Fonsc., with Asclepias for its
summer host and Nerium for the winter host.

136. Aphis oenotherae Oestlund

Oestlund, Minn. Geol. Nat. Hist. Surv., Bull. 4, p. 62, 1887 (orig. desc.;.
Clarke, Can. Ent., vol. 35, p. 252, 1903 (list).

Record. — Oenoihera bectiana; Epilobium sp., Berkeley (Clarke).

In 1903 Clarke recorded finding this species on primrose and
willow herb in Berkeley. Since then it has not been observed in Cali-
fornia. The author has had the opportunity to study specimens from
Minnesota, taken by A. C. Maxson.

A SYNOPSIS OF THE APHIDIDAE 119

137. Aphis oregonensis Wilson

Wilson, Trans. Am. Ent. Soc., vol. 41, p. 92, 1915 (orig. desc.).
Record. — Artemisia tridentata, California (Wilson).

Wilson stated to the author that he had taken this species in
California although he gave no locality or date records. On the
strength of his statement it is included among the California aphids.
The author has never seen specimens of it.

138. Aphis persicae-niger Smith

Figures 223, 224

Smith, Ent. Am., p. 101, 1890 (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 252, 1903 (list).

Gillette, Jour. Econ. Ent., vol. 1, p. 308, 1908 (desc.).

Weeks, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 244, 1912 (list).

Jones, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 318, 1912 (list).

Games, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 399, 1912 (list).

Wood, Mon. Bull. Cal. Comm. Hort., vol. 2, p. 570, 1913 (list).

Records. — Prunits spp. ; throughout California.

This species is ordinarily found infesting the tender twigs and
leaves of peach in the spring and early summer. Occasionally it is
found on nectarine, plum, and cherry. There are two records of its
occurrence on cherry known to the author; one in San Jose in May,
1912, by Davidson, and one in El Cajon, San Diego County, in May,
1916, by the author. Definite reports of its presence on peach come
from Los Angeles, Placer, Riverside, San Benito, San Bernardino,
San Diego, Santa Clara, and Tehama counties. In May, 1916, the
author observed it doing considerable damage to a young peach orchard
in the El Cajon Valley, San Diego County. Many of the twigs and
some of the larger branches were killed back for several inches, due
to the ravages of this insect.

The Hippodamia ladybird and the larvae of a syrphid fly were
abundant and devouring vast numbers of the aphids. However, it is
not often that this appears abundant enough to cause any great
amount of damage.

Its life history, although not thoroughly worked out, is interesting.
The following brief summary is from Essig:17

The insect winters over on the roots of the peach trees, where it may also be
found in the summer. The first aphids appear above ground very early in the

17 Essig, E. O., Beneficial and injurious insects of California; ed. 2. Suppl.
Mon. Bull. Cal. Comm. Hort., vol. 4, pp. 91-92, 1915.

120 MISCELLANEOUS STUDIES

spring and begin attacking the tender leaflets, shoots and suckers, usually those
at the base of the tree or nearest the ground. These first plant lice are all wing-
less. As soon as the buds, young fruit, and leaves appear they are promptly
attacked, the entire crop often being entirely ruined. The leaves are curled and
weakened, while the young fruit is so distorted as to be killed or rendered unfit
for market. During the months of April and May winged migratory females
appear, which start colonies on other trees. The work continues until about the
middle of July, when most of the lice leave the tops and again go to the roots.

139. Aphis pomi De Geer
Figures 225 to 227

De Geer, Memoires, vol. 3, p. 173, 1773 (orig. desc.).

Davidson, Jour. Econ. Ent., vol. 2, p. 301, 1909 (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1911 (list). Aphis mali Fabr.

Weatherby, Mon. Bull. Cal. Comm. Hort., vol. 1, p. 318, 1912 (list).

Games, Mon. Bull. Cal. Comin. Hort., vol. 1, p. 399, 1912 (list).

Branigan, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 285, 1915 (list).

Hurdley, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 445, 1915 (list).

Baker and Turner, Jour. Agr. Res., vol. 5, pp. 955-995, 1916 (complete

account).

Eecords. — Pyrus mains; Crataegus oxycantha; Catalpa sp. ; California.

In California this species has been reported on apple and haw-
thorn (Crataegus sp.) at Stanford University by Davidson and
Morrison; in Humboldt County by Weatherby; at Santa Rosa by
Games; and by others in Orange, Placer, Sonoma, Santa Cruz, San
Bernardino, and Monterey counties. Horticultural Commissioner
Armitage states that it has never been found in San Diego County,
and Horticultural Commissioner Norton writes that it is unknown in
Nevada County. These are the only two of the apple growing regions
of the state in which it is not known. The author has found it at
Stanford University on apple, catalpa, pear, and hawthorn, and at
Marysville on catalpa. Gillette lists loquat, quince, and flowering
crab as additional hosts. It seems to prefer the apple to other hosts,
and it is on the apple that its greatest injury is done. Gillette states :
' ' Among the apple trees it has its preference. Missouri Pippin seems
to be its first choice, while Rome Beauty, Black Twig, Ben Davis, and
a few others are second choice, and the Northern Spy is scarcely
attacked." The fact that the Northern Spy is almost immune is
interesting in that this variety is also quite immune to the devastations
of the woolly aphis (Eriosoma lanigera Hausman).

The life history of this aphid is quite similar to that of many other
species, and is as follows :

A SYNOPSIS OF THE APHIDIDAE 121

The eggs are laid in the fall of the year, probably during the latter
part of October, throughout November, and on into December. They
are laid for the most part on the smooth bark of the suckers and water
sprouts of the newer shoots. The author has found them in the
crotches of the twigs and stems where the bark is rougher, but this is
not the usual place. These eggs hatch in the spring about the time
the buds begin to show green. In California this is usually during
March, although some seasons it is as early as the middle of February,
depending entirely upon the weather conditions. These stem-mothers
at first feed on the young buds, until the latter have opened enough
to allow the aphids to crawl down into the curled leaves. Here they
feed for two or three weeks, when they mature and begin depositing
living young. This second generation consists chiefly of apterous
females, which mature in from two to four weeks and in turn produce
young. The following generations are in large part alate females
which migrate to other trees and there form new colonies. The alates
are most common at Stanford University during the latter part of
May and during the month of June. After June they seem to lessen
in number, perhaps due to the predaceous and parasitic enemies. The
first alates that the author has found in the spring were taken at
Stanford University on April 13, 1914. In the fall, often as early as
October, sexual males and females begin to appear, the males being
apterous, the females alate. These mate and very soon the female
lays its eggs. Egg laying begins usually in the latter part of October,
just as the leaves are beginning to fall, and continues into December
after the trees are bare. These eggs hatch in the spring into stem
mothers, and the life cycle is completed.

140. Aphis prunorum Dobr.

Figures 228 to 230

Dobrowljansky, Zur Biol. d. Blattlause d. Abstbaume u. Biirenstaucher, 1913

(orig. desc.).
Patch, Maine Agr. Exp. Sta., Bull. 233, p. 262, 1914 (desc. note).

Records. — Prunus domestica; Walnut Creek (Davidson) ; San Francisco, April,
1915 (Shinji).

A species of Aphis, supposed to be this species, has been taken on
prune and plum in the San Francisco Bay region. It agrees very well
with Dr. Patch's description listed. However, it may prove to be
synonymous with Siphocoryne nymphaeae (Linn.).

122 MISCELLANEOUS STUDIES

141. Aphis pseudobrassicae Davis

Figure 231
Davis, Can. Ent., vol. 46, p. 231, 1914 (orig. desc.).

Records. — Brassica spp. ; Walnut Creek (Davidson), San Diego, Riverside:
Baphanus sp., Riverside, September, 1916, June, 1917 : Matthiola annua, Riverside,
February to May, 1917.

Oftentimes in the spring this false cabbage aphis is found in large
colonies on radish, mustard, and so forth. Davidson has taken it in
the San Francisco Bay region, and the author throughout southern
California. The first few times that it was observed by the author
colonies of Aphis brassicae Linn, were also abundant. This led the
author to doubt its validity, and to undertake some breeding experi-
ments. In February, 1917, two colonies were started, each from one
alate female. They were followed through three generations, with the
result that all the individuals proved to be this species. At the same
time a colony of Aphis brassicae Linn, was started from one alate.
All the progeny of this individual proved to be the same. A. pseudo-
brassicae Davis differs from A. brass-icae Linn, in the following major
points :

A. pseudobrassicae Davis: A. brassicae Linn.:

Apterae not pulverulent. Apterae pulverulent.

Cornicles of apterae longer than hind Cornicles of apterae shorter than

tarsi. hind tarsi.

IV of alates with sensoria. IV of alates without sensoria.

142. Aphis ramona Swain

Figures 232 to 235
Swain, Trans. Am. Ent. Soc., vol. 44, p. 14, 1918 (orig. desc.).

Records. — Ramona stachyoides; Nordhoff and Santa Paula, Ventura County
(Swain).

This species has been taken twice in Ventura County by Essig.
It was described by the author from the specimens taken by Essig on
black sage.

143. Aphis rubiphila Patch

Patch, Maine Agr. Exp. Sta., Bull. 233, p. 269, 1914 (orig. desc.).
Records. — Rub us spp.; San Jose, May, 1916 (Davidson).

In the summer of 1916 Davidson found a species of Aphis infesting
loganberries and blackberries in San Jose, which was determined by
Dr. Patch as A. rubiphila Patch. Essig believes this to be a synonym

A SYNOPSIS OF THE APHIDIDAE 123

of A. gossypii Glover, but as the author has not had an opportunity
to study specimens he believes it best to recognize it as a distinct
species at present.

144. Aphis salicicola Thomas

Figures 188, 238, 237

Thomas, 111. Lab. Nat. Hist., Bull. 2, p. 8, 1879 (orig. desc.).
Williams, Univ. Neb. Studies, vol. 10, p. 139, 1910 (desc.).
Davidson, Jour. Econ. Ent., vol. 5, p. 408, 1912 (list).

Records. — Salix laevigata; Berkeley, June, 1915: Salix, sp. ; San Jose (David-
son).

This is an uncommon species, found in the San Francisco Bay
region on willow. The individuals are found in large colonies on
the terminal shoots and leaves. These colonies consist in large part
of apterae, there being but a very few alates. The species is quite
easily recognized by the long cornicles and by the very short second
branch of the third discoidal vein.

145. Aphis sambucifoliae Fitch

Figure 240

Fitch, Cat. Homop. N. Y., p. 66, 185 (orig. desc.).
Sanborn, Kan. Univ. Sci. Bull. 3, p. 52, 1904 (dese.).

Records. — Sambucus glauca; Oakland, April, 1915 (Essig) ; Berkeley, July,
1915.

In 1915 this species was taken twice, once by Essig in Oakland
and once by the author in Berkeley. This medium-sized black aphid
occurs in large colonies on the tender shoots and flower heads of the
common elderberry. In southern California the author has examined
hundreds of elderberry trees for this form, but has never found it.
Only once has he found any aphid on elderberry in the south, and
these proved to be Rhopalosiphum persioae (Sulz.).

146. Aphis senecio Swain

Figures 2, 4, 6, 241 to 245

Davidson, Jour. Econ. Ent., vol. 2, p. 302, 1909. Aphis sp. (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. A. bakeri Cowen (list).
Davidson, Jour. Econ. Ent., vol. 7, p. 133, 1914. A. bakeri Cowen (list).
Swain, Trans. Am. Ent. Soc., vol. 44, p. 16, 1918.

Records. — Abutilon sp. ; Stanford University, February, 1915: Ambrosia
psilostachya; Berkeley, 1915 (Essig): Amsinckia spp. ; Stanford University, 1909
(Davidson), 1912 (Morrison); Berkeley, 1915 (Essig): Anthemis spp.; San

124 MISCELLANEOUS STUDIES

Francisco Bay region, 1914 (Davidson) ; Pasadena, May, 1917 (Boy E. Camp-
bell): Artemisia spp.; San Francisco Bay region, 1914 (Davidson); Berkeley,
1915 (Essig) : Aster sp. ; San Diego, January, 1916; Ontario, January, 1917:
Bacchari* pilularis ; Berkeley, 1915 (Essig), Stanford University, 1916 (Ferris):
Calendula officinale; Berkeley, 1915 (Essig); San Diego, March, 1916; Riverside
and Orange, February, 1917: Chrysanthemum sp. ; Berkeley, 1914 (Essig); Octo-
ber, 1915; Menlo Park, San Mateo County, March, 1915; San Diego, January,
1916; La Jolla, February, 1916; Ontario, January, 1917: Cytisus proliferus;
Berkeley, 1915 (Essig): Gnapholium sp. ; Walnut Creek, 1914 (Davidson): Grin-
delia cuneifolia; Walnut Creek, 1915 (Davidson): Hclianthus annuus; San Frau-
cisco Bay region, 1914 (Davidson): Eumex sp. ; Stanford University, March,
1915: Salix sp. ; Berkeley, 1915 (Essig): Senecio spp.; Stanford University, 1909,
1910, 1914 (Davidson) ; Santa Paula, 1911 (Essig) ; Palo Alto, February, 1915.

This is a very common species throughout California, occurring
on many host plants, particularly the Compositae. It is found most
commonly in the early spring on asters, marigolds, and chrysanthe-
mums in southern California, and on German ivy and amsinckia in the
San Francisco Bay region. For sometime it was believed to be Aphis
bakeri Cowen, but its variety of host plants so widely different from
those of bakeri, led to its being identified as a distinct species. It is
one of the most common in the state, as a glance at the collection
records will show.

147. Aphis setariae Thomas

' Figures 246, 247

Thomas, 111. Lab. Nat. Hist., Bull. 2, p. 5, 1878 (orig. desc.).
Williams, Univ. Neb. Studies, vol. 10, p. 141, 1910 (desc.).

Record. — Prunus domestica; San Francisco Bay region (Davidson).

In some parts of the country this plum louse becomes abundant
enough to cause serious damage, but it has never been observed to be
so in California. Davidson writes that he has found it sparingly a few
times in the San Francisco Bay region. The author has never collected
it, but has had access to specimens from Morrison, taken in Indiana.

148. Aphis spiraecola Patch

Patch, Maine Agr. Exp. Sta,, Bull. 233, p. 270, 1914 (orig. desc.).

Records. — Spiraea spp.; Stanford University, 1912 (Morrison); Walnut Creek,
Contra Costa County, 1916 (Davidson).

In the San Francisco Bay region there is a small aphid very
similar to Aphis pami De Greer found attacking meadowsweet. David-
son and Morrison, who have both observed it, believe it to be this
species. The following brief descriptive notes are from alate females

A SYNOPSIS OF THE APHIDIDAE 125

taken by Dr. Patch on cultivated spiraea in Orono, Maine. These
notes are included here as there is no adequate description of this
species, the only ones18 being very meager notes indeed .

Alate viviparous females. — Body rather long and narrow, head
normal with no antennal tubercles. Antennae shorter than body,
reaching to about the base of the fourth abdominal segment. VI spur
the longest segment, followed by III, which is about two-thirds as
long. Following III are IV, V, and VI base. The usual primary
sensoria are present on V and VI, and the accessory sensoria on VI.
The secondary sensoria are fairly large and circular. There are six
or seven in an even line along the whole length of III. On IV there
may be one or two near the middle, or there may be none. Prominent
lateral tubercles are present on the prothorax and on the first and
seventh abdominal segments. The cornicles are fairly long, slender,
and taper slightly toward the apex. They are from one and one-half
to two times as long as the hind tarsi, and subequal to or very slightly
longer than the cauda. The cauda is fairly long, ensiform, slightly
constricted before the tip. The wings are normal, with the second
branch of the third discoidal nearer the apex of the wing than the
base of the first branch.

Measurements : Body length, 1.19 to 1.33 mm. ; width of thorax,
0.544 to 0.561 mm.; antennae total, 0.85 to 0.918 mm.; Ill, 0.17 to
0.1785 mm. ; IV, 0.136 to 0.153 mm. ; V, 0.1275 to 0.1445 mm. ; VI,
base 0.0935 to 0.102 mm. ; VI, spur 0.238 to 0.255 mm. ; cornicles, 0.1785
to 0.187 mm. ; cauda, 0.17 mm. ; hind tarsus, 0.102 mm. ; wing length,
1.97 to 2.04 mm.; width, 0.748 to 0.782 mm.; expansion, 4.55 mm.;
from base of first branch of third discoidal to wing tip, 0.578 to 0.68
mm. ; from base of second branch to wing top, 0.17 to 0.255 mm.

149. Aphis tetrapteralis Cockerell

Cockerell, 'South. Cal. Acad. Sci., Bull. 1, p. 4, 1902 (orig. desc.).
Record. — Atriplex canescens tetraptera; La Jolla (Cockerell).

This species has been observed but once, when described by
Cockerell. He writes: "It differs from Aphis atriplices Linn, by its
smaller size, mode of life, and shorter cornicles. It seems to be
related to Aphis vnonardae Oestlund." In 1916 the author spent
considerable time hunting for this species in the vicinity of La Jolla,
but in vain.

is Patch, Edith M., Maine Aphids of the Rose Family. Maine Agr. Exp. Sta.,
Bull. 233, p. 270, 1914, Aphis spiraecola n.n.; Gillette, C. P., Plant louse notes,
Family Aphididae. Jour. Econ. Ent., vol. 3, p. 404, 1910. Aphis spiraeella Schout.

126 MISCELLANEOUS STUDIES

150. Aphis viburnicolens n.sp.

Becords. — Viburnum tinua; Riverside, February to May, 1917; Redlands, Feb-
ruary, 1917; Orange, February, 1917: Laurus rotoundifolia, Riverside, March,
1917.

In the early spring there is a small green and black aphid that
attacks in great numbers the racemes of laurustinus and laurel in
Southern California. In fact, it is so abundant at times as to seriously
injure the plants by preventing them from flowering. The leaves
and buds are very sticky and covered with the sooty mold fungus.
During April, 1917, all the aphids left the laurel and laurustinus, but
the alternate host has as yet not been observed. Specimens were sent
to Gillette and Patch for determination, but neither could identify
them. Dr. Patch wrote as follows:

This insect is not spiraecola, a slide of which I am sending you.

spiraecola sp.

Cornicles longer than III Cornicles shorter than III

VI spur longer than III VI spur subequal to III

VI spur longer than IV and V VI spur subequal to IV and V

IV subequal to V IV longer than V

I do not know this species. I do not have spiraeella Schout. for comparison.

Gillette stated concerning this species: "This is a species of Aphis
close to, but almost certainly distinct from, spiraeella Schout., and so
far as we know, may be new. ' '

From this it would appear that the species from laurustinus and
laurel is a new species, and it is described herewith as such.19 Cotype
specimens are in the author's private collection, in the collection of
the University of California in Berkeley, and of the Citrus Experi-
ment Station in Riverside.

Alate viviparous female. — Prevailing color green. Head and
thorax dusky brown to black. Antennae dusky to black. Beak light
brown with tip black. Tibiae, femora of fore legs, and basal one-half
of femora of middle and hind legs brown ; tarsi, tips of tibiae, tips of
fore femora, and apical one-half of middle and hind femora black.
Abdomen pale to apple green, sometimes with a few dusky marginal
spots. Cornicles and cauda black.

is The species reported by Davidson (Jour. Econ. Ent., vol. 3, p. 377, 1910)
as Aphis mali Fabr. from Lauras laurustinus (Viburnum tinus?) and by Essig
(Injurious and Beneficial Insects of California, Mon. Bull. Cal. Comm. Hort.,
Supp. vol. 4, p. xlvi, 1915) as Aphis pomi De Geer from laurustinus, are probably
this species.

A SYNOPSIS OF THE APHIDIDAE 127

Head normal, with frontal and antennal tubercles absent. An-
tennae short, reaching only to the second abdominal segment. Ill
and VI spur subequal ; IV and V subequal and about three-fourths as
long as III or VI spur. The usual primary sensoria are present on V
and VI, and the accessory sensoria on VI. Secondary sensoria are
found on III and IV, from five to nine on the former and from one
to four on the latter. Cornicles short, subcylindrical, and tapering
from base toward apex. Cauda fairly long, ensiform, with a slight
constriction in the middle, the cauda is slightly longer than the hind
tarsi, and the cornicles a little longer than the cauda. Lateral tuber-
cles are present on the prothorax, and on the first, fourth, and seventh
abdominal segments. The cornicles are subequal to IV or V. The
hind tarsi are somewhat longer than VI base. The wings are fairly
large, with regular venation, the second joint of the third discoidal
arising about half way between the tip of the wing and the base of
the first joint.

Measurements: Body length, 1.214 to 1.479 mm. (av. 1.372 mm.) ;
width of thorax, 0.476 to 0.578 mm. (av. 0.5338 mm.) ; antennae total,
0.733 to 0.918 mm. (av. 0.8925 mm.) ; HI, 0.187 to 0.230 mm. (av.
0.2067 mm.) ; IV, 0.136 to 0.161 mm. (av. 0.1473 mm.) ; V, 0.119 to
0.153 mm. (av. 0.1416 mm.) ; VI, base 0.085 to 0.102 mm. (av. 0.0877
mm.) ; VI, spur 0.204 to 0.230 mm. (av. 0.216 mm.) ; cornicles, 0.127
to 0.153 mm. (av. 0.1422 mm.) ; cauda, 0.110 to 0.136 mm. (av. 0.1252
mm.) ; hind tarsi, 0.102 to 0.119 mm. (av. 0.1023 mm.) ; wing length,
1.921 to 2.397 mm. (av. 2.167 mm.) ; width, 0.799 to 0.935 mm. (av.
0.8704 mm.) ; expansion, 4.42 to 5.304 mm. (av. 4.875 mm.).

Apterous viviparous female. — General color green with the follow-
ing dusky to black: head, antennae, apex of beak, cornicles, cauda,
distal margin anal plate, tarsi, and tips of tibiae. Legs, except tarsi
and tips of tibiae, dusky brownish green. Antennae reach to the base
of the second abdominal segment. The various segments are propor-
tionally the same as in the alates. The beak reaches to the distal
margin of the first coxae or almost to the apical margin of the third
coxae. Lateral body tubercles are present on the prothorax and first,
second, and seventh abdominal segments. Sometimes they are also
present on the third, fourth, or fifth abdominal segments as well. The
cornicles and cauda are subequal, each slightly longer than the hind
tarsi, and of the same form as in the alates.

Measurements: Body length, 1.326 to 1.462 mm. (av. 1.3685 mm.) ;
width of thorax, 0.595 to 0.68 mm. (av. 0.6975 mm.) ; antennae total

128 MISCELLANEOUS STUDIES

0.731 to 0.782 mm. (av. 0.748 mm.) ; III, 0.153 to 0.187 mm. (av.
0.170 mm.) ; IV, 0.119 to 0.136 mm. (av. 0.1224 mm.) ; V, 0.119 mm.;
VI, base 0.085 mm.; VI, spur 0.136 to 0.1995 mm. (av. 0.1632 mm.) ;
cornicles, 0.153 to 0.1995 mm. (av. 0.170 mm.) ; cauda, 0.136 to 0.170
mm. (av. 0.162 mm.) ; hind tarsi, 0.102 to 0.119 mm. (av. 0.114 mm.).

151. Aphis yuccae Co wen

Figures 303 to 305

Cowen, Colo. Agr. Exp. Sta., Bull. 31, p. 122, 1895 (orig. dese.).
Williams, Univ. Neb. Studies, vol. 10, p. 145, 1910. Aphis yuccicola n.sp.
(desc.).

Records. — Yucca moJuivensis; Moorpark, Ventura County, April, 1916 (F. M.
Trimble) ; San Diego, May, 1916.

In April, 1916, Horticultural Inspector F. M. Trimble of Ventura
County sent the author a few specimens of the alate and apterous
viviparous females of this species, taken on Spanish dagger in Moor-
park. In the latter part of the next month the author found a few
apterae on the leaves of Spanish dagger in Golden Hill Park, San
Diego. There were only a few individuals present at that time, but
there was evidence of an earlier heavy infestation. Following are a
few notes to supplement Williams ' excellent description of this species.

Ill is the longest segment of the antennae, followed by VI spur,
which is about three-fourths as long. IV is next, being a little over
one-half as long as III and about five-sixths as long as VI spur. V
is slightly shorter than IV and is followed closely by VI base, which
is about one-half the length of the spur. The usual primary sensoria
are present on V and VI and the accessory sensoria on VI (fig. 303).
The apterae have no secondary sensoria, while the alates along the
whole length of III (fig. 304) have about twenty-five irregularly
placed sensoria of irregular size. VI is without sensoria. Lateral
tubercles are present on the prothorax and on the first and seventh
abdominal segments. The cornicles (fig. 305) are long and slightly
tapering, being but slightly shorter than the spur of the sixth antennal
segment and about twice as long as the hind tarsi. The cauda (fig.
305) is ensiform or sickle-shaped and about three-fourths as long as
the cornicles. In length it is about equal to the fifth antennal seg-
ment and one-half again as long as the hind tarsi.

A SYNOPSIS OF THE APHIDIDAE

129

Alate viviparous females. — Measurments : Body length, 1.78 to 1.9
mm. (av. 1.86 mm.) ; width, (thorax), 0.95 mm.; antennae total, 1.38
to 1.51 mm. (av. 1.449 mm.) ; III, 0.34 to 0.425 mm. (av. 0.391 mm.) ;
IV, 0.238 to 0.273 mm. (av. 0.256 mm.) ; V, 0.212 to 0.229 mm. (av.
0.219 mm.) ; VI, 0.136 to 0.17 mm. (av. 0.155 mm.) ; spur, 0.255 to
0.306 mm. (av. 0.289 mm.) ; cornicles, 0.255 to 0.2975 mm. (av. 0.275
mm.) ; cauda, 0.2125 to 0.238 mm. (av. 0.225 mm.) ; hind tarsi, 0.153
mm.; wing length, 3.06 to 3.4 mm. (av. 3.19 mm.) ; wing width, 1.27
to 1.46 mm. (av. 1.338 mm.) ; wing expansion, 7.48 mm.

30. Genus Toxoptera Koch.
Koch, Die Pflanzenlause, p. 253, 1857. Type Aphis aurantii Fonsc.

152. Toxoptera aurantii (Fonsc.)

Figures 114, 163, 276

Boyer de Fonscolombe, Ann. Ent. Soc. France, vol. 10, 1841. Aphis (orig.

desc.).

Essig, Pom. Jour. Ent., vol. 3, p. 601, 1911. T. aurantiae Koch (desc.).
Davis, U. S. Dept. Agr., Bur. Ent., Tech. Ser., Bull. 25, pt. 1, p. 8, 1912.

Records. — Citrus spp. ; throughout citrus sections of southern and central Cali-
fornia (Essig, author) ; San Jose (Davidson).

This is the common black louse of the citrus trees, and is found at
almost any time of the year on the younger and more tender leaves
of various species of Citrus. It is more or less heavily preyed upon
by the braconid fly, Lysiphlcbus testaceipcs Cresson. In fact, the
author has noticed several infestations in which fully ninety-five per
cent of the individuals were parasitized. Besides these the syrphid
flies cause great havoc among colonies. Of these the author has reared
Allograpta obliqua Say from a colony taken in the vicinity of El
Cajon, San Diego County. Never does this species become abundant
enough to seriously damage trees, due undoubtedly to the effective
work of its predacious and parasitic enemies. Only in the spring
are they found to any great extent, although occasionally throughout
the year small infestation can be noticed.

130 MISCELLANEOUS STUDIES

31. Genus Hyalopterus Koch

Koch, Die Pflanzenlause, p. 17, 1854. Type Aphis arundinis Fabricius ( A.
pruni Fabr.).

153. Hyalopterus arundinis (Fabr.)

Figures 181, 185, 186

Fabricius, Ent. Syst., vol. 4, p. 212, 1749. Aphis (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 247, 1903 (list).

Davidson, Jour. Econ. Ent., vol. 2, p. 303, 1909 (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910 (list).

Essig, Mon. Bull. Cal. Comm. Hort., vol. 2, p. 569, 1913 (list).

Essig, Mon. Bull. Cal. Comm. Hort., vol. 3, p. 624, 1913. A. prunifoliae

Fitch (list).

Weldon, Mon. Bull. Cal. Comm. Hort., vol. 2, p. 630, 1913 (list).
Weldon, Mon. Bull. Cal. Comm. Hort., vol. 3, p. 378, 1914 (list).
Patch, Maine Agr. Exp. Sta., Bull. 233, 266, 1914 (dese.).
Davidson, Mon. Bull. Cal. Comm. Hort., vol. 6, p. 64, 1917 (note).

Records. — Prunus spp., Phalaria, arundinacea, Phragmites communis, Typha
latifolia; central California.

During the spring and early summer of the year this ' ' mealy-plum
louse" is often very abundant on various species of Prunus in the
central part of the state, especially in the San Francisco Bay region
and the Sacramento Valley. As summer continues all the aphids
desert the plum for other host plants, where they remain until fall.
The summer hosts in California so far known are reed grass, canary
grass, and tule, or cat-tail rush. In the Santa Clara Valley there is a
feeling among the prune growers that this aphid is the cause of the
splitting of the prunes, which is often quite extensive. However, this
remains to be proven.

32. Genus Liosomaphis Walker
Walker, The Zoologist, p. 1119, 1868. Type Aphis berberidis Kalt.

154. Liosomaphis berberidis (Ealt.)

Figures 184, 251, 252

Kaltenbach, Monog. d. Pflanzenlause, p. 85, 1843. Aphis (orig. desc.).
Davis, Ann. Ent. Soc. Am., vol. 1, p. 254, 1908. Bhopalosiphum (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 378, 1910. Bhopalosiphum (list).

Eecords. — Berberis vulgaris; Stanford University (Davidson); February to
May, 1915; Berkeley, June to August, 1915.

This species is found throughout the year on the lower sides of
the leaves of barberry in the San Francisco Bay region. The apterae
are often very abundant, but the alates are always quite scarce. This

A SYNOPSIS OF THE APHIDIDAE 131

species is similar to species of Rhopalosiphum, particularly in the
shape of the cornicles and cauda, but owing to the absence of antennal
tubercles it falls into the tribe Aphidini instead of Macrosiphini.
Hence Walker's genus Liosomaphis is maintained for this species.

33. Genus Siphocoryne Passerini

Passerini, Gli Afidi, 1860. Type Aphis pastinacae Linn, (xylostei Schrank).

There has been much diversity of opinion concerning this genus,
some aphidologists considering it as Siphocoryne Passerini, some as
Hyadaphis Kirkaldy, and some as a synonym of Rhopalosiphum Koch.
This last is incorrect as this is most certainly not a Macrosiphini for
the antennal tubercles are lacking. In 1904 Kirkaldy proposed the
name Hyadaphis to replace Siphocoryne, but in the author's opinion
this is uncalled for, so he maintains the original name, Siphocoryne
Passerini.

There have been reported from various parts of California eight
species of Siphocoryne as follows: capreae (Fabr.), conii (Dvdn.),
foeniculi (Schrank), nymphaeae (Linn.), pastinacae (Linn.), salicis
(Monell), umbellulariae (Dvdn.), and xylostei (Schrank). There
are, however, really but three species; capreae (Fabr.), nymphaeae
(Linn.) and pastinacae (Linn.). According to Gillette,20 S. salicis
Monell is a synonym of S. capreae (Fabr.), and xylostei (Schr.) of
pastinacae (Linn.). Davidson21 states that S. conii (Dvdn.) is a
synonym of xylostei ( Schr. ) , and therefore it is the same as pastinacae
(Linn.). Morrison writes that the specimens Davidson called S. foeni-
culi (Schr.) are capreae (Fabr.), and those he described as Hyadaphis
umbellulariae n.sp. are S. pastinacae (Linn.). These two species,
pastinacae (Linn.) and capreae (Fabr.), have been greatly confused
but Gillette22 has worked out their synonymy quite satisfactorily. The
following key for distinguishing them is from his paper.

Joints 4, 5, 6, and antennal spur subequal, the spur usually distinctly the
longest, cornicles fully three-fourths as long as third joint of the antenna, a small
tubercle on the alate form and a large one on the apterous individuals always
present capreae

20 Gillette, C. P., Two Rhopalosiphum species and Aphis pulverulens n.sp.,
Jour. Econ. Ent., vol. 4, pp. 320-325, 1911.

21 Davidson, W. M., Plant louse notes from California, Jour. Econ. Ent., vol. 7,
p. 133, 1914.

22 Gillette, C. P., Two Rhopalosiphum species and Aphis pulverulens n.sp.,
Jour. Econ. Ent., vol. 4, pp. 320-325, 1911.

132 MISCELLANEOUS STUDIES

Joint 623 of the antenna distinctly shorter than 5, the fourth still shorter
and its spur nearly as long as joints 4, 5, and 6 combined, cornicles seldom much
exceeding one-half the third joint of the antenna in length, and a supra-caudal
tubercle or spine entirely absent pastinacae

Aphis nymphaeae Linn, has usually been considered by American
aphidologists as a species of Rhopalosiphum, but the presence of lat-
eral body tubercles, the short, robust body, and the absence of antennal
tubercles place it in the Aphidini rather than the Macrosiphini.
Therefore, it must be considered as belonging to this genus. Baker24
has recently recognized it as belonging here.

KEY TO CALIFORNIAN SPECIES

1. A small spine or tubercle present at the distal end of the body just above the

cauda (figs. 255, 256) capreae (Fabr.)

— No supra-caudal tubercle or spine 2

2. General color pale green. VI spur as long as IV, V and VI base combined.

Cornicles at most but slightly more than one-half the length of III.

pastinacae (Linn.)

— General color dark brown, wine, or black. VI spur not as long as IV, V and

VI base combined, although longer than any two together. Cornicles and
III subequal nymphaeae (Linn.)

155. Siphocoryne capreae Fabr.

Fabricius, Ent. Syst., p. 211, 1794. Aphis (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 252, 1903. S. foeniculi (Pass.) (list).
Davidson, Jour. Econ., vol. 2, p. 303, 1909. S. solids Monell (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. S. foeniculi (Pass.), (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. S. salicis Monell (list).
Essig, Pom. Jour. Ent., vol. 3, p. 534, 1911. Hyadaphis pastin-acae (Linn.)
(desc.).

Becords. — Foeniculum vulgare; Berkeley and Newcastle (Clarke), Stanford
University (Davidson): Carum spp. ; Cicuta virosa; Santa Paula, Berkeley
(Essig) : Salix laevigata; Santa Paula (Essig), Brea Canyon, Los Angeles County,
April, 1917; Riverside, May, 1917:- Salix nigra; Lakeside, San Diego County,
April, 1916: Salix sp., Stanford University (Davidson).

This species is found more or less abundantly in the spring on the
tender shoots and leaves of willows, migrating in early . summer to
various species of Umbelliferae. It is more common than 8. pastinacae
(Linn.), which species is also found on Umbelliferae in the summer,
but which passes the fall, winter, and spring on honeysuckle.

23 In all the author 's specimens, VI is shorter than V, which in turn is shorter
than IV, while VI spur is nearly as long as the three together.

2* Baker, A. C. and Quaintanee, A. L. Aphids injurious to orchard fruits,
currant, gooseberry and grape, U. S. Dept. Agr., Farmers' Bulletin 804, p. 21,
1917.

A SYNOPSIS OF THE APHIDIDAE 133

156. Siphocoryne nymphaeae Linn.

Figure 172

Linnaeus, Syst. Nat., vol. 2, p. 734, 1735. Aphis (orig. dese.).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. Rhopalosiplium (list).
Essig, Pom. Jour. Ent., vol. 4, p. 793, 1912. Ehopalosiphum (desc.).
Davidson, Mon. Bull. Cal. Comm. Hort., vol. 6, p. 65, 1917. Ehopalosiphum
(note).

Records. — Polygonum sp., Alisma sp., Potamogeton sp. ; San Francisco Bay
region (Davidson) : Typha latifolia; Santa Paula (Essig), San Francisco Bay
region (Davidson): Nymphaea sp.; San Francisco Bay region (Davidson),
Fresno, June, 1915: Prunus domestica; Berkeley, 1916 (Essig).

This aphid occurs throughout the summer months on various semi-
aquatic plants, lily, tule, and so forth. In the fall it migrates to
plum, where eggs are laid. The first two or three generations in the
spring occur on plum, but about June there is a migration to its sum-
mer host plants. So far it has been found in southern California
only in Ventura County.

The species listed as Aphis prunorum Dobr. (see no. 140) may be
this species. Essig believes it is, but the author is not certain so does
not list it as a synonym.

157. Siphocoryne pastinacae Linn.

Figures 266 to 270

Linnaeus, Syst. Nat., p. 451, 1735. Aphis (orig. desc.).
Davidson, Jour. Eeon. Ent., vol. 2, p. 304, 1909. S. xylostei (Schr.) (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 304, 1909. S. conii n.sp. (desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 377, 1910. S. xylostei (Schr.) and

S. conii Dvdn. (list).
Davidson, Jour. Econ. Ent., vol. 4, p. 599, 1911. Hyadaphis umbellulariae

n.sp. (desc.).

Davidson, Pom. Jour. Ent., vol. 3, p. 399, 1911. S. conii Dvdn. (list).
Davidson, Jour. Econ. Ent., vol. 7, p. 133, 1914. S. xylostei (Schr.) (list).

Records. — Lonicera sp. ; Stanford University (Davidson), Claremont (Essig),
Berkeley, April, 1915: Umbellularia calif ornica; San Jose (Davidson): Conium
maculatum; Stanford University, Penryn, Placer County, and San Jose (David-
son).

This aphid occurs on honeysuckle during the winter and spring,
and on various semiaquatic plants in the summer. It has been taken
in southern California, in the San Francisco Bay region, and in the
Sacramento Valley.

134 MISCELLANEOUS STUDIES

34. Genus Myzaphis Van der Goot

Van der Goot, Ziir Systematik der Aphiden, Tijdscrift voor Entomologie,
vol. 56, p. 96, 1913. Type Aphis rosarum Walker.

The author believes that this genus of Van der Goot's should be
accepted for the two following species: Aphis abietina Walker and
Aphis rosarum Walker. A. rosarum has usually been considered as
belonging to the genus Myzus, but the absence of antennal tubercles
excludes it from that genus (see figs. 306-308, 313). The cornicles
and cauda are not typical of Aphis, and these together with the dis-
tinctive frontal tubercle on the head and the absence of lateral body
tubercles distinguish it from Aphis. Consequently this genus should
be recognized. Following is a key for separating the two known
species, both of which occur in California :

Cornicles slightly clavate (figs. 312, 315), shorter than III. Ill tuberculate, IV

without sensoria (fig. 309). Found on Rosa spp rosarum (Walker)

Cornicles cylindrical (fig. 197), equal to or longer than III. Ill with 9 to 12
rather large secondary sensoria, IV with 1 to 4 (fig. 196). On conifers.

abietina (Walker)

158. Myzaphis abietina (Walker)

Figures 196, 197

Walker, Ann. Mag. Nat. Hist., vol. 3, p. 301, 1848. Aphis (orig. desc.).
Wilson, Proc. Ent. Soc. Brit. Columbia, June, 1915 (desc.).
Record. — Picea excelsa; San Francisco, March, 1915 (Compere).

The only report of this species in America is that of Wilson, who
found it on spruce (Picea sp.) at Vancouver, British Columbia. On
March 26, 1915, Harold Compere of San Francisco took a number of
specimens of this species on the twigs of Norway spruce (Picea
excelsa) in Golden Gate Park, San Francisco. The specimens are in
Essig 's and the author's collections.

159. Myzaphis rosarum (Walker)

Figures 308 to 317

Walker, Ann. Mag. Nat. Hist., voL 3, 1848. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 3, p. 379, 1910. Myzus (list)
Eecords. — Rosa spp.; Stanford University (Davidson); Santa Paula (Essig),
San Diego, March to July, 1916.

This species has been reported in the San Francisco Bay region
by Davidson and in Santa Paula by Essig. In the Bay region it is
rather scarce and is second to Macrosiphum rosae (Linn.) in abun-
dance on roses. The author has taken it at Stanford University in
1915, and in San Diego several times in 1916. In San Diego in 1916
it was by far the most abundant rose-infesting aphid. The author

135

has observed it in such numbers on roses as to cover the undersides
of practically all the leaves and the calyx cups of the flowers. In some
cases the buds were stunted and the flowers unshapely from its effect.
In the rose garden of the Panama-California International Exposition
these aphids were of considerable importance, necessitating continual
care to keep them under control.

Since there is no adequate description of this species in the Ameri-
can aphid literature the author describes it herewith. The following
description was drawn from ten specimens of alate and eight of
apterae, collected in Santa Paula, Stanford University, and San Diego.

Alate viviparous female. — Color notes (taken from notes made
at the time of collection of specimens at Stanford in March, 1915) :
Head, antennae, and thoracic plates black. Abdomen pale apple green
with smoky blotch on dorsum. Legs : apical two-thirds of femora
smoky, basal one-third pale, tibiae pale except dusky tip, tarsi dusky.
Cornicles green (dusky), cauda pale apple green.

Head is twice as wide as long with a fairly distinct tubercle on the
front (fig. 308). Antennal tubercles are lacking or very indistinct.
Antennae reach almost to the base of the third abdominal segment
(figs. 309, 310). Ill is the longest segment, followed by IV, spur, V,
and VI. The spur and IV are practically equal. Of sixteen antennae
examined, in three, the spur and IV were equal, in ten, IV was slightly
longer than the spur, while in three, the spur was slightly longer than

IV. V is slightly shorter than the spur, and VI slightly shorter than

V. However, IV, spur, V, and VI are all almost equal. On V and VI
are the usual primary sensoria, and VI the accessory sensoria (fig.
300). Ill is tuberculate, being furnished with a large number of
irregularly placed secondary sensoria (fig. 309). IV is without any
sensoria. The beak reaches almost to the second coxae.

The prothorax is without lateral tubercles. The wings are normal,
being about twice the length of the body. The second branch of the
cubitus arises nearer the apex of the wing than the base of the first
branch (fig. 311). In but one of seventeen specimens examined was
the origin of the second branch of the cubitus nearer the base of the
first branch than the tip of the wing. In this specimen the measure-
ments were : 0.561 mm. from tip of wing to base of first branch and
0.289 mm. from tip of wing to base of second branch.

The abdomen is long and narrow and is without lateral body
tubercles. The cornicles (fig. 312) are long, being but slightly shorter
than the third antennal segment, and over twice as long as the hind
tarsi. They are slightly clavate on the inner side. The cauda (fig.

136 MISCELLANEOUS STUDIES

312) is long and pointed (ensiform), being slightly more than one-half
as long as the cornicles and about one-half as long again as the hind
tarsi.

Measurements : Body length, 1.19 to 1.41 mm. (av. 1.28 mm.);
width of thorax, 0.459 to 0.527 mm. (av. 0.487 mm.) ; antennae total,
0.85 to 1.156 mm. (av. 1.027 mm.) ; III, 0.255 to 0.34 mm. (av. 0.317
mm.) ; IV, 0.1275 to 0.2295 mm. (av. 0.1768 mm.) ; V, 0.119 to 0.17
mm. (av. 0.1365 mm.) ; VI, 0.085 to 0.119 mm. (av. 0.1095 mm.) ;
spur, 0.119 to 0.204 mm. (av. 0.1695 mm.) ; cornicles, 0.238 to 0.306
mm. (av. 0.2574 mm.) ; cauda, 0.136 to 0.187 mm. (av. 0.1588 mm.) ;
hind tarsi, 0.119 to 0.136 mm. (av. 0.1205 mm.) ; wing length, 2.482
to 2.72 mm. (av. 2.5483 mm.) ; wing width, 0.884 to 1.02 mm. (av.
0.9396 mm.) ; wing expansion 5.423 to 5.967 mm. (av. 5.5836 mm.).
From tip of wing to base of first branch of cubitus 0.561 to 1.037 mm.
(av. 0.8041 mm.) ; from tip of wing to base of second branch of cubitus,
0.17 to 0.34 mm. (av. 0.2907 mm.).

Apterous viviparous female. — Head about as long as broad with a
large prominent tubercle on the front, this tubercle being considerably
larger than in the alate form ; in some individuals it is fully as large
as the first antennal segment (fig. 313). Antennal tubercles small but
distinct, similar to those of the alate. Antennae (fig. 314) short,
reaching only to the third coxae. Ill is the longest segment, followed
by the spur, IV, VI, and V. These are all subequal, the formula of
the averages being spur, IV, VI, and V. The formulae for seven
antennae are S, VI (V, IV) ; S, (VI, V, IV) ; S, V, IV, VI; S, IV (V,
VI) ; S, (IV, V), VI; IV (V, VI, S) ; (S, VI, IV), V. The usual
primary sensoria are present, but there are no secondary sensoria.
The beak is short, reaching only to the second coxae.

The prothorax is without tubercles. The thorax is normal, as are
the legs. The abdomen is long and narrow, without lateral tubercles,
and without long capitate hairs as found in some species of Myzus.
The cornicles (fig. 315) are long, cylindrical, and slightly tapering
toward the apex, or slightly clavate at apex. They are over twice as
long as the third antennal segment and over three times as long as
the hind tarsi (fig. 317), and half as long again as the cauda. The
cauda (fig. 316) is long and ensiform, being slightly more than twice
the length of the hind tarsi, and about two-thirds the length of the
cornicles.

Measurements: Body length, 1.275 to 1.615 mm. (av. 1.428 mm.) ;
width of thorax, 0.493 to 0.748 mm. (av. 0.6375 mm.) ; antennae total,
0.544 to 0.731 mm. (av. 0.6239 mm.) ; III, 0.153 to 0.238 mm. (av.

A SYNOPSIS OF THE APHIDIDAE 137

0.178 mm.) ; IV, 0.068 to 0.119 mm. (av. 0.0855 mm.) ; V, 0.068 to
0.102 mm. (av. 0.0833 mm.) ; VI, 0.068 to 0.119 mm. (av. 0.085 mm.) ;
spur, 0.085 to 0.136 mm. (av. 0.117 mm.) ; cornicles, 0.306 to 0.442
mm. (av. 0.3655 mm.) ; cauda, 0.204 to 0.272 mm. (av. 0.2338 mm.) ;
hind tarsi, 0.102 mm. (Note : no color notes were taken of the apterae
at the time of collection and as all the specimens were killed in alcohol,
dehydrated in xylene and mounted in Canadian balsam, it is impossible
to give any color notes.)

35. Genus Coloradoa Wilson

Wilson, Ann. Ent. Soc. Am., vol. 3, p. 323, 1910. Type Aphis rufomaculata
Wilson.

This genus was described by Wilson in 1910 to contain the species
Aphis rufomaculata Wilson. After examining specimens of this
species recently, the author is of the opinion that Coloradoa and
Myzaphis are synonymous, for there does not seem to be enough differ-
ence between this species and the two species of Myzaphis to warrant
a separation of genera. However, the author does not feel certain
concerning the point, so lists both these genera. Should they later
prove to be synonymous, Myzaphis would have to be dropped and
replaced by Coloradoa. There is but one species belonging to this
genus.

160. Coloradoa rufomaculata Wilson

Wilson, Ent. News, vol. 14, p. 261, 1908. Aphis (orig. desc.).
Eecord. — Chrysanthemum, cultivated; Sacramento, April, 1917 (Davidson).

The author has recently received specimens of this species from
Davidson taken on chrysanthemum in Sacramento.

36. Genus Cerosipha Del Guercio

Del Guercio, Nouve relazione agraria di Firenze, vol. 2, p. 116, 1909. Type
C. passeriniana n.sp.

161. Cerosipha cupressi Swain

Swain, Trans. Am. Ent. Soc., vol. 44, p. 19, 1918 (orig. desc.).
Records. — Cupressus guadelupensis ; San Diego, 1916; Riverside, 1917; C.
macrocarpa, San Diego, 1916.

This species, recently described by the author, has been taken by
him several times in San Diego and Riverside on blue cypress and
Monterey cypress. It is an extremely interesting little aphid, differ-
ing considerably from any other species known to the author, both
in habits and appearance. Its five-jointed antennae, long cauda,
atrophied cornicles, and convexity of abdomen are quite distinctive.

138 MISCELLANEOUS STUDIES

Subfamily Pemphiginae Mordwilko25

Mordwilko, Ann. Mus. Zool. Imp. Acad. Sci. St. Petersburg, vol. 13, pp.
362-364, 1908.

A summary of Mordwilko 's description of this subfamily has
already been given. The latest and probably the most complete sys-
tematic work on this subfamily that has been done is that of Dr. Albert
Tullgren of Stockholm, Sweden, in his paper, ' ' Aphidologische
Studien I" in 1909. Tullgren divides this subfamily into six tribes,
viz: Vacunina, Hormaphidina, Mindarina, Pemphigina, Schizoneu-
rina, and Anoeciina. In the tribe Vacunina he places Vacuna Heyden
and Glyphina Koch ; in Hormaphidina is the one genus Hamamelistes
Shimmer; in Mindarina is the one genus Mindarus Koch; in Pem-
phigina he places Asiphum Koch, Pachypappa Koch, ProciphUus
Koch, Thecdbius Koch, and Pemphigus Hartig; in Schizoneurina he
places the two genera, Schizoneura Hartig, and Tetraneura Hartig;
and finally in the Anoeciina is found the one genus Anoecia Koch. It
can be seen that he uses several of Koch 's genera which have not here-
tofore been generally used, namely: Prociphilus Koch, Thecabius
Koch, Asiphum Koch, and so forth. Lately there has been a tendency
among American aphidologists to accept these genera, and thus to
divide up the larger genus Pemphigus into these smaller ones. Mord-
wilko in his keys divides this subfamily into four groups, namely:
Hormaphidina, Pemphigina, Schizoneurina, and Vacunina. In Hor-
maphidina he includes besides the genus Hamamilestes Shimmer, the
genera Hormaphis Osten-Sacken and Cerataphis Lichtenstein. In
Pemphigina he includes Pentaphis Heyden, Tetraneura Hartig,
Pemphigus Hartig, Aploneura Passerini, Rhizoctonus Horvath, and
Paracletus Heyden. In Schizoneurina he places Lowia Lichtenstein,
Colopha Monell, Pachypappa Koch, Schizoneura Hartig, Anoecia, Koch,
and Mindarus Koch. In Vacunina he includes but the one genus
Vacuna Heyden, which he does not separate from Glyphina Koch.

There is considerable difference in the classifications of these two
authors, but as far as we are concerned here in California our genera
are placed about the same by both. Following is a translation of
Mordwilko 's key to the groups :

2", The author has under way a more exhaustive study of this subfamily, par-
ticularly of the species of Pemphigus and Procipliilus. As this research is still
in progress, however, it was thought best to omit any report of it, the author here
confining himself merely to the records of the presence of the various species in
California. It was hoped to have this study completed at the present time, but
the unprecedented conditions of this season have made it necessary to delay further
study for the time being.

A SYNOPSIS OF THE APHIDIDAE 139

1. Winged forms with a cucurbit-shaped cauda. Nymphs that failed to molt

with three-jointed antennae. Winged forms with three to five-jointed
antennae, which are coarsely ringed from the third on. Wingless partheno-
genetic females presenting the appearance of the larvae of other families,
as of some kinds of Coccidae, or of species of Aleyrodes. Sexual forms
with beaks Group Hormaphidina

— Winged forms without distinct cauda. Nymphs that failed to molt with four

to five- jointed antennae. Antennae of winged females five-to six- jointed.
Sensoria may be found on the third and following joints, often in the
form of arches or half rings, but never as complete rings 2

2. Cubitus [third discoidal vein] of the fore wings simple. Cornicles, which are

pore or pointlike, present only in some species, and then not in all forms.

Group Pemphigina

— Cubitus [third discoidal vein] of fore wings once-branched. Cornicles mostly

point or pore-like 3

3. Antennae of winged forms six-jointed. Wings held roof-like when at rest.

Group Schizoneurina

— Antennae of winged forms five-jointed. Wings held flat when at rest.

Group Vacunina

Group Hormaphidina Mordw.

Mordwilko, Ann. Mus. Zool. Imp. Acad. Sci. St. Petersburg, vol. 13, pp.
364-365, 1908.

The antennae of the winged forms are 5- to 3-jointed (?). With the exception
of the first two joints they are closely and entirely ringed. Even in the genus
Hormaphis O.-S., where the antennae are 3-jointed, they may probably be con-
sidered morphologically as of five joints. The wings are held flat at rest. There
are four transverse veins on the fore wings, the third of which [third discoidal]
is simple. The first two [first and second discoidals] originate at the same point
on the subcosta. The hind wings have one or two transverse veins, in the latter
case both originating at the same point. The wingless parthenogenetic females
on the alternate host plants (for example on birch) are mostly circular in shape,
and have small wax tubes around them. Other forms are coccid-like. The sexual
forms have beaks. The cornicles are absent.

This is a description as given by Mordwilko in the above mentioned
paper. Below is a key to the genera, as given by Mordwilko and by
Van der Goot, the latter of whom includes in this group the two
genera Hamamelestes Shimmer and Ceratophis Licht.

1. Antennae of winged females plainly five-jointed 2

— Antennae of winged females only three-jointed Hormaphis O.-S.

2. Antennae always five-jointed. Front of head always with two little horns.

Third discoidal once-branched Cerataphis Lichtenstein

— Antennae of apterous forms three- or four-jointed. Front without horns.

Third discoidal simple Hamamelistes Shim.

140 MISCELLANEOUS STUDIES

37. Genus Cerataphis Liechtenstein

Lichtenstein, Bull. Soci6te ent. de France, vol. 2, p. 16, 1882. Type Coccus
lataniae Boisd.

162. Cerataphis lataniae Boisduval

Boisduval, Ent. Hort., 1867. Coccus (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 5, p. 404, 1912 (list).
Essig, Univ. Calif. Publ. Entom., vol. 1, p. 342, 1917 (list).
Eecords. — Fern, Stanford University (Davidson) ; orchid, Oakland (Essig).
This coccid-like species has been reported twice in the San Fran-
cisco Bay region, by Davidson and by Essig. Morrison and the author
have also taken it on the same ferns on which Davidson found it in
the Stanford University nursery.

Group Pemphigina Lichtenstein

Below is a key to the California genera of this group, adapted
from Mordwilko, Tullgren and Del Guercio. Del Guercio described
a genus in 1909 for Pemphigus radicicola Essig, which he called
Trifidaphis.

1. Antennae of alate females five-jointed Trifidaphis Del Guer

— Antennae of alate females six-jointed 2

2. Stem mothers with five- join ted antennae. Wax-gland plates on head always

present and usually large. Spring and fall migrants with wax-gland plates
always on mesothorax and abdomen, and usually on head. Dorsal pores
never present 3

— Stem mothers with four -jointed antennae. Head normally without wax-gland

plates. Dorsal pores sometimes present. Stem mothers and spring migrants
(fundatrix and fundatrigenia) at first live in the same closed galls.

Pemphigus Hartig

3. Secondary sensoria furnished with hairy fringe (Wimperkranz). Wax-gland

plates generally large. In stem mothers there appear four very large pro-
notal wax-gland plates, placed in a transverse row. All plates have a
clearly chitinized border. Stem mother and migrants live together.

Prociphilus Koch

— Secondary sensoria without hairy fringe (Wimperkranz). Wax-gland plates

generally small. In stem mothers there are six pronotal plates, of which
the four middle ones are arranged in the form of a trapezium. In the
winged fall migrants (sexupara) there are also transverse abdominal gland
plates, which are without clearly chitinized borders. Stem mothers and
spring migrants live in separate galls Thecabius Koch

A SYNOPSIS OF THE APHIDIDAE 141

38. Genus Trifidaphis Del Guercio

Del Guercio, Eiv. di patal. veg., vol. 3, p. 20, 1909. Type Pemphigus radi-
cicola Essig.

163. Trifidaphis radicicola Essig

Essig, Pom. Jour. Ent., vol. 1, p. 8, 1909. Pemphigus (orig. desc.).
Baker, Pom. Jour. Ent., vol. 1, p. 74, 1909. (Translation of Del Guer-
cio's description of the genus.)
Essig, Pom. Jour. Ent., vol. 2, p. 283, 1910 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912 (list).

Eecords. — Amaranthus retroflexus, Solanum douglasii; Claremont, Santa Paula
(Essig).

Essig described this species from specimens taken on the roots of
Amaranthus retroflexus and Solanum douglasii in Santa Paula and
Claremont. Later Del Guercio described a new genus for this species
based on the venation and the antennae. It seems that the type speci-
men of this species had but five-jointed antennae and so of course
it could not belong to the genus Pemphigus. On an examination of
eight specimens, including the type specimen and seven cotypes, the
author finds that the number of joints in the antennae are variable.
The type specimens had both antennae with but five joints. Six
antennae had but five joints, six had six distinct joints, and four had
five joints in which the division into six could be made out. This
divison was in the third joint at about one-third the distance from the
apex. Consequently one could say that this species was typically five-
jointed, but with some specimens with the third joint divided into two,
or it could be said that it was typically six-jointed, but in some speci-
mens a reducton occurred through the joining of the third and fourth
segments. As but a few specimens were examined the author is not
willing to state which is the more common, hence leaves this as a valid
genus, although he is of the opinion that this really belongs to the
genus Prociphilus Koch.

39. Genus Pemphigus Hartig

Hartig, Jahresb. u. d. Fortschr. d. Forstwiss. u. forstliche Naturk., vol. 1,
p. 645, 1837. Type Aphis bursarius Linn.

This genus is represented in California by three well known
species,26 P. betae Doane, P. populi-caulis Fitch, and P. populi-trans-

26 There has been taken several times a species forming elongate leaf galls on
Populus fremontii, both in the San Francisco Bay region by Davidson and in San
Diego County by the author, that structurally seems to be identical with P. populi-
caulis Fitch, but its gall is quite distinct, being more or less similar to that of
P. betae Doane. Further study may reveal the identity of this form.

142 MISCELLANEOUS STUDIES

versus Riley. All of these species, during at least a part of their life
cycles, infest various species of Populus, where they form more or less
distinctive galls.

KEY TO FUNDATRIGENIAE27

1. Secondary sensoria present only on III. Galls formed on leaf petioles, with a

transverse opening on the outside of the curve populi-transversus Eiley

— Secondary sensoria on other segments as well as on III 2

2. Secondary sensoria on III to VI inclusive. Galls formed by the twisting of

the petiole with an oblique opening on the inside of the curve.

populi-caulis Fitch

— Secondary sensoria on III and IV.2® Gall formed on the under side of the

leaves, being more or less elongate and opening on the upper side.

betae Doane

164. Pemphigus betae Doane

Doane, Ent. News., vol. 11, p. 390, 1900 (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 248, 1903 (list).

Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909 (list).

Davidson, Jour. Ecou. Ent., vol. 3, p. 372, 1910 (list).

Williams, Univ. Neb. Studies, vol. 10, p. 92, 1910. P. balsamiferae n.sp.

(desc. fundatrigenia).

Essig, Pom. Jour. Ent., vol. 4, p. 299, 1912 (list).
Maxson, Jour. Econ. Ent., vol. 9, p. 500, 1916 (note).

Becords. — Beta vulgaris; San Francisco Bay region, Monterey County, Sacra-
mento Valley. (Rumex spp., Chenopodium spp., etc.?)

Under the name P. betae Doane, Clarke, Davidson, and Essig have
reported a species of aphid infesting the roots of sugar beets, dock,
Chenopodium, and other plants throughout California.

Originally this species was described from specimens taken on
sugar beet in Washington, but later29 it was proven that a species
forming elongated leaf galls on Populus balsamifera in the spring
migrated to beets, and was identical with this species. In 1916 Maxson
(cited above) states that his investigations point to the fact that in
Colorado there are more than one species of Pemphigus attacking the
sugar beet, one of which is this species that forms the elongate leaf
gall on poplar in the spring, and which is known now as P. betae
Doane.

27 At present only a key to the alate migrants or fundatrigeniae occurring in
galls on poplar is given. It is hoped that later, keys to all forms may be formu-
lated. At present, however, the life histories of the species are not sufficiently
known.

28 The sexupara or alate migrants from beets to poplars have secondary sen-
soria on III to V inclusive. These form no galls on poplar, however.

29 Parker, The life history of the sugar-beet root louse, Jour. Econ. Ent., vol.
7, pp. 136-141, 1914;

Gillette, Notes on some Colorado aphids having alternate host plants, Jour.
Econ. Ent., vol. 8, p. 97, 1915.

A SYNOPSIS OF THE APHIDIDAE 143

These observations of Maxson's together with those made by the
author lead to the conclusion that all the reported cases of infestation
of beets and other hosts by P. betae Doane in California do not neces-
sarily refer to this species. Never have the fuiidatrix or fundatrigenia
been taken on poplar in California. This strengthens the point that
the aphids on beets and other hosts may not all be P. betae Doane.
Further studies and observations will have to be made before this
point can be settled, however.

165. Pemphigus populicaulis Fitch

Fitch, Eep. Ins. N. Y., vol. 5, p. 845, 1859 (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 248, 1903 (list).

Davidson, Jour Econ. Ent., vol. 2, p. 299, 1909 (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910 (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910. P. populi-transversus Biley

(list).
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. P. populi-transversus Riley

(list).

Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912 (list).
Essig, Pom. Jour. Ent., vol. 4, p. 708, 1912 (dese.).
Davidson, Jour. Econ. Ent., vol. 8, p. 420, 1915 (sexuales).

Records. — Populus fremontii, P. trichocarpa; from Placer County to San Diego
County (Clarke, Davidson, Essig, Morrison, and the author).

The species infests cottonwoods throughout the state, forming a
gall by the twisting of the leaf petiole. The sexuales are found,
according to Davidson, under the bark where the eggs are also laid.
The author has found the species in San Diego County, having taken
the fundatrix, virgogenia, and fundatrigenia in galls in May, 1916,
and the dead sexupara at the same time in old galls. These latter
probably died without ever leaving the galls.

166. Pemphigus populi-transversus Riley

Eiley, U. S. Geog. Geol. Surv., Bull. 5, p. 15, 1880 (orig. desc.).
Essig, Univ. Calif. Publ. Entom., vol. 1, p. 343, 1917 (list).

Records. — Populus fremontii, Berkeley, September, 1914 (Essig), Eiverside
September to October, 1916, May to July, 1917.

This species forms large galls on the leaf petioles of poplar some-
what similar to the preceding species, differing in that the opening is
on the opposite side of the gall, and is transverse rather than oblique.
Essig 's specimens were determined by Gillette, the author's by Max-
son. Davidson reported a species under this name from Stanford

144 MISCELLANEOUS STUDIES

University, but later wrote the author that he was mistaken in his
determination, the species being P. populicaulis Fitch instead.

Just recently the author received specimens of the sexupara of this
species from J. R. Parker, Bozeman, Montana. These were taken by
S. H. Jones in Port Allen, Louisiana, in September, 1915, on the roots
of cabbages. Jones notes that cabbage and other cruciferous plants
are the alternate host of this species. This spring the author received
a large number of apterae of a species of Pemphigus taken in Orange
County on the roots of cabbage. A specific determination of the
species was impossible but it may have been this one.

40. Genus Thecabius Koch
Koch, Die Pflanzenlause, p. 294, 1857. Type Pemphigus affinis Kalt.

This genus is very similar to Prociphilus, and by some authors,
particularly Baker,30 is considered as synonymous. However, for
present purposes the author proposes to retain it for the three species
included herewith.

KEY TO CALIFOBNIAN SPECIES

1. Antennae short, barely reaching to the metathorax, and not one-third as long

as the body. Ill but slightly longer than VI populi-monilis Eiley

— Antennae longer, reaching beyond the base of the abdomen, and about one-half

as long as the body. Ill considerably longer than VI 2

2. V and VI with secondary seusoria populi-conduplifolius Cowen

— VI without secondary sensoria californicus Davidson

167. Thecabius californicus (Davidson)

Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910. Pemphigus ranunculi n.sp.
(orig. desc.).

Davidson, Jour. Econ. Ent., vol. 4, p. 414, 1911, renamed Pemphigus cali-
fornicus Dvdn.

Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912. Pemphigus (desc. ala. and
apt. female).

Davidson, Jour. Econ. Ent., vol. 7, p. 127, 1914 (note).

Records. — Ranunculus californicus; San Francisco Bay region (Davidson, Mor-
rison, Essig, author): ? Populus sp. ; Walnut Creek, Contra Costa County, May,
1915 (Davidson) : Fraxinus oregona; Walnut Creek (Davidson).

This aphid is found quite abundantly on the roots and stems of
the small California buttercup in the San Francisco Bay region.
According to Davidson there is a migration during April from butter-

so Baker, A. C., Identity of Eriosome pyri, Jour. Agr. Ees., vol. 5, p. 1118,

A SYNOPSIS OF THE APHIDIDAE 145

cup to ash. There may be a migration to poplar as well, for the
.author has specimens that seem to be this species taken by Davidson
•on poplar. Gillette31 places this species as a synonym of T. populi-
•conduplifolius Cowen, which attacks both Ranunculus and Populus
in Colorado. Davidson, however, is convinced that they are distinct.

168. Thecabius populiconduplifolius (Cowen)

Cowen, Colo. Agr. Exp. Sta., Bull. 31, p. 115, 1895. Pemphigus (orig.

desc.).

Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. Pemphigus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912. Pemphigus (list).
Gillette, Annals Ent. Soc. Am., vol. 7, p. 61, 1914 (desc. and life history).

Record. — Populus trichocarpa; Stanford University (Davidson).

This species was reported by Davidson on poplar at Stanford
University. Since then no further records of its occurrence in the
state have been made. In Colorado, Gillette finds that the common
buttercup, Ranunculus sp., is an alternate host and so considers the
preceding species as a synonym. This may be possible, but it is quite
doubtful.

169. Thecabius populimonilis (Riley)

Eiley, U. S. Geol. Surv., Bull. 5, p. 13, 1879. Pemphigus (orig. dese.).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. Pemphigus (list)
Davidson, Pom. Jour. Ent., vol. 3, p. 398, 1911. Pemphigus (list).
Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912. Pemphigus (list).
Gillette, Ann. Ent. Soc. Am., vol. 6, p. 485, 1913 (desc. and life history).

Records. — Populus spp. ; Tulare and Placer counties (Davidson); Santa Paula
(Essig), Eiverside, 1916-1917.

Throughout central and southern California this species is found
on various species of Populus where it forms more or less globular
galls on the upper side of the leaves near the margins. In the vicinity
of Riverside the young stem mothers began to appear in April (1917).
When first observed in September, 1916, nearly all the galls were
empty while a few contained alate migrants (sexupara probably).
According to Gillette the eggs are laid on the trunks of Populus, thus
the entire life cycle is passed on the one host plant. This is rather
unusual for the Pemphiginae of this section.

si Gillette, C. P., Some Pemphiginae attacking species of Populus in Colorado,
Ann. Ent. Soc. Am., vol. 7, pp. 61-65, 1914.

146 MISCELLANEOUS STUDIES

4] . Genus Prociphilus Koch

Koch, Die Pflanzenlause, p. 279, 1857. Type Aphis bumeliae Schrank.
KEY TO CALIFORNIAN SPECIES

1. Stigma of forewings conspicuously darkened. V with a few annular secondary
sensoria, VI with or without any. Dorsal thoracic wax plates small and

oval alnlfoliae (Williams)

— Stigma not conspicuously darkened. V and VI without annular secondary
sensoria. Dorsal thoracic wax plates quite large and triangular.

venafuscus Patch

170. Prociphilus alnifoliae (Williams)

Williams, Univ. Neb. Studies, vol. 10, p. 91, 1910. Pemphigus (orig. desc.).
Baker, Jour. Agr. Ees., vol. 5, p. 1118, 1916 (note).

Records. — Heteromeles arbutifoliae ; Sespe, Ventura County, March, 1915 (S.
H. Essig); May, 1915 (C. P. Clausen).

There has been no record of this species from California heretofore,
but the author has specimens taken on California holly or Christmas
berry in Sespe Canyon during March and May, 1915, by S. H. Essig
and C. P. Clausen.

171. Prociphilus venafuscus Patch

Patch, Ent News, vol. 20, p. 319, 1909. Pemphigus (orig. desc.).

Essig, Pom. Jour. Ent., vol. 3, p. 553, 1911. Pemphigus fraxini-dipetalae
n.sp. (orig. desc.).

Essig, Pom. Jour. Ent., vol. 4, p. 699, 1912. Pemphigus fraxini-dipetalae
Essig (list).

Childs, Mon. Bull. Cal. Comm. Hort., vol. 3, p. 220, 1914. Pemphigus
fraxini-dipetalae Essig (list).

Wilson, Trans. Am. Ent. Soc., vol. 41, p. 85, 1915. Prociphilus fraxini-
dipetalae (Essig) (note).

Davidson, Jour. Econ. Ent., vol. 8, p. 421, 1915. Prociphilus fraxini-
dipetalae (Essig) (list).

Baker, Jour. Agr. Kes., vol. 6, pp. 1118-1119, 1916 (desc. notes, synonymy).

Records. — Fraxinus dipetala; Santa Paula (Essig), Contra Costa and Santa
Clara counties (Davidson): F. oregona; Oregon (Wilson); Berkeley, April, 1915:
Aesoulus calif or nicus ; Sacramento (Childs): Pseudotsuga taxifolia; Oregon (Wil-
son).

Occasionally this very large aphid is found infesting the leaves
of ash in the San Francisco Bay region and in the mountains of
southern California. In early summer it leaves the ash, and according
to Wilson infests the roots of Douglas fir in Oregon. At one time
Leroy Childs found a few specimens on buckeye in the vicinity of
Sacramento, but it is probable that these were accidental there.

A SYNOPSIS OF TEE APE I DID AE 147

Group Schizoneurina Lichtenstein

This group as considered by Mordwilko contains the following
genera : Lowia Licht., Colopha Monell, Pachypappa Koch, Schizoneura
Hartig, Anoecia Koch, and Mindarus Koch. Tullgren places in his
tribe Schizoneurina the two genera, Schizoneura Hartig, and Tetra-
neura Hartig. Pachypappa Koch he places in his tribe Pemphigina,
and he has a separate tribe for each of the genera Anoecia Koch and
Mindarus Koch, calling them respectively tribe Anoeciina and tribe
Mindarina. Below is a translation of Mordwilko 's key.

1. Wings laid flat on back when at rest Lbwia Licht.

— Wings held roof -like when at rest 2

2. Stigma of forewings trapezoidal in shape, reaching only to the beginning of

the curve around the end of the wing, never extending to the tip of the
wing. Radial vein originating from the posterior exterior corner of the
stigma 3

— Stigma linear, very long, reaching to the wing tip on the front side of the

wing, and even following the backward curve of the exterior side of the
wing to some extent. Eadial vein starting almost at the beginning to the
interior edge of the stigma. Sexual forms with beaks Mindarus Koch

3. Hind wings with one transverse vein Colopha Monell

— Hind wings with two transverse veins 4

4. Both transverse veins originating from the same point on the longitudinal

veins Pachypappa Koch

. — Transverse veins of hind wings originating separately 5

5. Bodies of apterous and alate forms with little hair, and covered at least on the

dorsum of the abdomen with waxy powder. Cornicles pore-like (point-like).
Sexual forms without beaks Eriosoma Leach

— Bodies of apterous and alate forms very hairy and not covered with waxy

powder or granules (only the stem mothers are weakly pulverulent). Cor-
nicles comparatively large, tuberculate (cone-like). Sexual forms with
beaks '. Anoecia Koch

The genera Lowia Licht., Pachypappa Koch, and Anoecia Koch
are not represented in California. Colopha Monell and Mindarus
Koch are both represented by their type species. It has been proven
that Eriosoma Leach has priority over Schizoneura, Hartig, so that
genus is now known by that name. It is represented in California by
three or four species at present.

148 MISCELLANEOUS STUDIES

42. Genus Colopha Monell

Monell, Can. Ent., vol. 9, p. 102, 1877. Type Byrsocrypta ulmicola Fitch.

172. Colopha ulmicola (Pitch)

Fitch, Rept. Ins. N. Y., vol. 4, p. 63, 1858. Byrsocrypta (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909 (list).
Patch, Maine Agr. Exp. Sta., Bull. 181, 196, 1910 (desc.).

Record. — Ulmus sp.; Stanford University (Davidson).

Davidson recorded this species from elm at Stanford University in
1909. Since then it has not been found again.

43. Genus Eriosoma Leach

Leach, Trans. Hort. Soc. London, vol. 3, p. 54, 1820. Type Aphis lani-
gerum Hausman.

Until quite recently this genus has been known as Schizoneura
Hartig, but as Baker32 has pointed out, the name Eriosoma, has
priority. In California there are three distinct species represented,
with a possible fourth. One of these is known only on elm, one on
apple (and elm), and one on pear (and elm).

The following key to the fall migrants is adapted partially from a
table of Baker and Davidson.33

1. Body naked except caudal segment. Distal sensoria of V and VI with fringe.

languinosa (Hartig)

— Body with some woolly covering. Distal sensoria without fringe 2

2. Wing veins narrow without brown margins. Ill longer than IV, V, and VI

together lanigerum (Haus.)

— Wing veins broad with brownish margins. Ill not so long as IV, V, and VI.

americana (Biley)

173. Eriosoma americana (Riley)

Eiley, TJ. S. Geol. Surv., Bull. 5, p. 4, 1879. Schizoneura (orig. desc.).
Clarke, Can. Ent., vol. 35, p. 248, 1903. Schizoneura (list).
Patch, Maine Agr. Exp. Sta., Bull. 220, p. 268, 1913. Schizoneura (desc.
note).

Becords. — Ulmus americana; Berkeley (Clarke); Walnut Creek, June, 1915
(Davidson) ; Palo Alto, May, 1915.

This leaf-curling aphid of the American elm is found in. the San
Francisco Bay region, and in some cases is very abundant. In May

32 Baker, A. C., The woolly apple aphis, U. S. Dept. Agr., Office Sec 'y, Report
101, pp. 11-12, 1915.

sa Baker, A. C., and Davidson, W. M., Woolly pear aphis, Jour. Agr. Ees.,
vol. 6, p. 358, 1916.

A SYNOPSIS OF THE APE I Dl DAE 149

and June, 1915, it was especially so on a row of elms on the campus
of Stanford University. At that time stem mothers, nymphs, and
alate spring migrants were present in the galls. By the last of June
all of these had flown away, leaving the galls empty. According to
Baker elm is the only host plant of this species.

174. Eriosoma lanigerum (Hausman)

Hausman, Mag; Ins., vol. 1, p. 440, 1802. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909. Schizoneura (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 374, 1910. Scliizoneura (list).
Baker, U. S. Dept. Agr., Office Sec'y, Report 101, pp. 11-16, 1915 (desc.
and biology).

Record. — Pyrus mains, throughout the state.

Wherever apple trees are found in the state this woolly aphis is
also found ; the white masses on the trunks and leaves being very con-
spicuous, the colonies on the roots more injurious but less conspicuous.
In California only the -apple has been found to be attacked. The
winter is passed by young nymphs on the roots. As the warmer
weather of spring comes these migrate up the trunks and out on
the branches and twigs. Here they feed throughout the summer. In
the fall there is a downward migration, and occasionally a fall
migrant is seen. Whether or not these fly to elms as in other parts of
the country, is not known, but none have ever been observed on elm.

175. Eriosoma languinosa (Hartig)

Hartig, Zeitschr. Ent., vol. 3, p. 359, 1841. Aphis (orig. desc.).

Baker and Davidson, Jour. Agr. Res., vol. 6, pp. 351-360, 1916. E. pyricola

n.sp. (desc.).
Baker and Davidson, Jour. Agr. Res., vol. 10, pp. 65-74, 1917. E. pyricola

B. & D. (desc. and biology).

Records. — Pyrus communig, TJlmus campestris; central California.

In 1916 Baker and Davidson described a species of Eriosoma that
attacks the roots of pears throughout the central part of the state,
naming it E. pyricola. Later Davidson found that a species common
on Ulmus campestris was the alternate form of this species. This elm
form checks up very favorably with specimens of E. languinosa Hartig
from Europe, and is undoubtedly identical. Thus the name pyricola
will have to be dropped in favor of languinosa. These elm galls are
of a rather peculiar shape, and, as Patch writes, they have the appear-
ance of a bonnet.

150 MISCELLANEOUS STUDIES

44. Genus Mindarus Koch
Koch, Die Pflanzenlause, p. 277, 1857. Type M. abietinus n.sp.

176. Mindarus abietinus Koch

Koch, Die Pflanzenlause, p. 278, 1857 (orig. desc.).

Clarke, Can. Ent., vol. 35, p. 248, 1903. Scliizoneura panwola Thos. (list).

Patch, Maine Agr. F.xp. Sta., Bull. 182, p. 242, 1910 (desc.).

Eecords. — Pinu-s radiata; Berkeley, Palo Alto (Clarke): Abies cilicia; Stan-
ford University, May, 1915.

This aphid, easily recognized by the extremely long stigma of the
fore wings, has been found in the San Francisco Bay region infesting
the shoots of Monterey pine and Cilician fir.

Group Vacunina Mordwilko

This group contains but two genera, Vacuna Hey den and Glyphina
Koch. Mordwilko does not recognize Glyphina, as distinct from
Vacuna, although Tullgren does. The latter separates the two genera
as follows:

1. Last abdominal tergite formed into a knob-shaped tail. Integument bare,

and at most partially set with short lancet-shaped hairs Vacuna Heyd.

— Last abdominal tergite half-moon shaped, strongly swollen, but scarcely,
if at all, separated from the base. Integument set with stiff bristle-like
hairs and in apterous females with grain-like elevations ....Glyphina Koch3*

45. Genus Vacuna Hey den

Heyden, Ent. Beitr., vol. 2, p. 289, 1837. Type Aphis dryophila Schrank.

177. Vacuna dryophila (Schrank) (?)

Schrank, Fauna Boica, vol. 1, p. 113, 1801. Aphis (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 7, p. 128, 1914. Chaitophorus sp. (desc.).
Davidson, Jour. Econ. Ent., vol. 10, p. 290, 1917 (desc.).

Eecord. — Quercus lobata; Walnut Creek (Davidson).

Recently Davidson described this species from specimens taken on
valley oak in Contra Costa County, where he had observed it for three
years. The single alate female he has taken does not appear identical
with European specimens of V. dryophila, so he lists the species under
this name provisionally.

This genus is not represented in California.

A SYNOPSIS OF TEE APHIDIDAE

151

Subfamily Phylloxerinae Dreyfus

This subfamily consists of two groups, the Chermisina and the
Phylloxerina. Below is a key to these two groups taken from Van
der Goot:

1. Body always with wax glands. Antennae of adults three-jointed, seemingly
five-jointed, with three large sensoria. Gonapophyses appearing as three

short lips. Sexuales dwarfed, with beak Group Chermisina

— Body usually without wax glands. Antennae of adults three-jointed, with two
large sensoria. Gonapophyses seem to be lacking. Sexuales dwarfed, with-
out beak Group Phylloxerina

Group Chermisina Borner

This group consists of three genera, Pineus Shimmer, Cnapholodes
Macq., and Chermes Linn, as it is generally considered, although some
authors add more, as Gillettea Del Guercio and Guercioja Mordw. In
California but one of these genera is represented, and that by but
two species.

46. Genus Chermes Linnaeus
Linnaeus, Syst. Nat., vol. 10, 1758. Type Chermes sambuci Linn.

178. Chermes cooleyi Gillette

Gillette, Proe. Acad. Nat. Sci. Phila., vol. 69, p. 3, 1907 (orig. desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909. C. coweni Gill. (list).
Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910. C. coweni Gill. (list).
Brannigan, Mon. Bull. Cal. Comm. Hort., vol. 4, p. 285, 1915 (list).

Records. — Pseudotsuga taxi folia, Pinus pinea; San Francisco Bay region, Sac-
ramento Valley.

This species was first reported in California by Davidson, who
found it on Douglas fir at Stanford University. Essig lists it from
San Francisco, San Mateo, and Santa Clara counties on Douglas fir.
In 1915 it was reported twice, once in Sacramento on Douglas fir, and
once on Italian stone pine. The author has specimens from E. J.
Vosler taken in Sacramento where it was found infesting the twigs
and needles of Italian stone pine. Only the apterous females were
present, however.

152 MISCELLANEOUS STUDIES

179. Chermes pinicorticis Fitch

Fitch, Trans. N. Y. State Agr. Soc., vol. 14, p. 971, 1855. Coccus (orig.

desc.).

Storment, 20th Ann. Eep. Illinois St. Ent., appendix, 1898 (desc.).
Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909 (list).
Davidson, Jour. Eeon. Ent., vol. 3, p. 372, 1910 (list).

Record. — Pinus pinaster maritima; Stanford University (Davidson).

This species, which is unknown to the author, was reported as
present at Stanford University on Pinus pinaster maritima, where
it was so abundant as to sometimes kill the young trees. For a com-
plete description see Storment 's paper listed above.

Group Phylloxerina Borner

There are two genera in this tribe, as considered by Borner and
Mordwilko, although the American authors have generally taken cog-
nizance of but one, namely, Phylloxera Boyer. Below is a key from
Mordwilko to these genera.

1. Neither wingless females nor any other forms secreting any waxy material.

Phylloxera Boyer
— Wingless females secreting a waxy powder Phylloxerina Borner

47. Genus Phylloxera Boyer

Boyer de Fonscolmbe, Ann. Ent. Soc. France, vol. 3, p. 222, 1834. Type
P. quercus Boyer.

180. Phylloxera vitifoliae Fitch

Fitch, Eept. Ins. N. Y., vol. 1, p. 58, 1855 (orig. desc.).

Planchon, C.-E. Aead. Sci. Paris, vol. 67, pp. 588-594, 1868. P. vastatrix

(desc.).

Clarke, Can. Ent., vol. 35, p. 248, 1903. P. vastatrix Plan (list).
Davidson, Jour. Econ. Ent., vol. 2, p. 299, 1909. P. vastatrix Plan. (list).

Davidson, Jour. Econ. Ent., vol. 3, p. 372, 1910. P. vastatrix Plan. (list).
Records. — Grape; Central and Northern California.

This is the only species of this genus reported in California. It
is one of the most destructive species of plant lice in this section of
the country, having in its time practically wiped out the grape indus-
try of Santa Clara Valley, and of many other parts of the state. It
seems that in California this species infests the roots only of the grape,
the forms that produce the leaf galls in the eastern parts of the
country not being found here.

A SYNOPSIS OF THE APHIDIDAE 153

48. Genus Phylloxerina Borner

Borner, Arbeiter aus d. kais. biol. Anst. f. Land- und Forstwirtschaft,
vol. 6, pp. i-v, 81-320, 1908. Type Phylloxera salicis Linn.

This genus is represented in California by two species, one found
on the stems of cotton wood (Populus sp.) and the other on the stems
and exposed roots of willow (Salix sp.).

181. Phylloxerina popularia (Pergande)

Pergrande, Proc. Davenport Acad. Sci., vol. 9, p. 266, 1904. Phylloxera

(orig. desc.).
Davidson, Jour. Econ. Ent., vol. 8, p. 420, 1915. Phylloxera (list).

Records. — Populus spp. ; Walnut Creek (Davidson), Merced (Beers).

The only report of this species in California is the one of Davidson
who found it on Populus fremonti and Populus trichocarpa at Walnut
Creek. On October 14, 1915, A. A. Beers of Merced sent some speci-
mens to the author from balm of Gilead (Populus balsamifera) in
Merced. These were all apterous females, and were found in great
masses of white wax on the smaller branches and twigs. These reports
are the only ones since its original report from Texas and Louisiana
by Pergande.

182. Phylloxrina salicola (Pergande)

Pergande, Proe. Davenport Acad. Sci., vol. 9, p. 267, 1904. Phylloxera

(orig. desc.).
Davidson, Jour. Econ. Ent., vol. 8, p. 419, 1915. Phylloxera (list).

Records. — Salix spp.; Walnut Creek (Davidson); Pasadena (Smith).

This species was also reported from Walnut Creek by Davidson
on arroyo willow (Salix lasiolepis) where he found it on the stems
and exposed roots. On October 13, 1915, A. G. Smith sent the author
specimens from an ornamental willow (Salix sp.) in Pasadena, where
he found it very abundantly that fall. The specimens were all
apterous females, and were found in the midst of considerable masses
of wax. This species has only been reported from Illinois, District
of Columbia, and California.

154 MISCELLANEOUS STUDIES

APPENDIX 1
KEYS TO THE GENERA AND TRIBES OF APHIDIDAE

BY
P. VAN DEE GOOT, 1913

Subfamily APHIDINAE

1. Antennae seven-jointed (better six-jointed). The last true joint with a dis-

tinct, more delicate continuation (terminal process). This continuation
almost as long as, or even much longer than the last segment; if shorter,
the cauda is distinctly wart-shaped, and the number of rudimentary gona-
pophyses is always two. Cornicles almost always well formed and clearly
projecting. Wings with twice-branched cubitus, only once-branched in
exceptional cases 2

— Antennae mostly six-jointed, the last joint with a short projection, this being

usually distinctly shorter than half the last segment. Cornicles scarcely
projecting, very often only appearing as pores or entirely absent. Wings
with a simple or once-branched cubitus 5

2. Cauda wart-like, occasionally not so, or scarcely separated, but then the number

of rudimentary gonapophyses is always distinctly two 3

— Cauda sickle-shaped or knobbed, not wart-like, only very seldom absent. Rudi-

mentary gonapophyses always three Siphonophorina

3. Cornicles very long, almost cylindrical. Rudimentary gonapophyses three.

Drepanosiphina

— Cornicles very short, somewhat clubbed. Rudimentary gonapophyses two or

four 4

4. Number of rudimentary gonapophyses four. Body never with long clubbed

hairs Chaitophorina

— Number of rudimentary gonapophyses two. Body often with knobbed hairs.

Tarsi always with two pulvillae [Haftlappchen] Callipterina

5. Cauda wart-like 6

— Cauda not wart-like, usually absent 7

6. Anal plate bilobed. Sensoria of alate females linear Hormaphidlna

— Anal plate simple. Sensoria of alate females circular Vacunina p.p.

7. Cauda distinctly sickle-shaped Mindarina

— Cauda only scarcely or not at all separated 8

8. Antennae five-jointed. Cornicles very short, only slightly projecting. Body

without distinct wax gland groups Vacunina p.p.

— Antennae six-jointed, those of the apterous forms often only four- or five-

jointed. Cornicles often only pores or entirely lacking. Body often with
wax gland 9

9. Body with long, mostly fine hairs; without distinctly facetted wax gland

plates. Primary sensoria almost always, without hairy edges 10

— Body naked; very often with distinctly facetted wax-gland plates. Primary

sensoria often with hairy edges 11

10. Rudimentary gonapophyses three. Wings mostly with twice-branched cubitus.
Cornicles always prominent Lachnina

— Rudimentary gonapophyses none. Wings with simple or once-branched cubitus.

Cornicles often absent ... ....Anoeciina

A SYNOPSIS OF THE APHIDIDAE

155

11. Eudimentary gonapophyses three. Facets of wax-gland plates almost equal-
sized. Wings with simple cubitus. Sensoria of alate forms long oval, not
linear Pemphlgina

— Eudimentary gonapophyses none. Wax gland plates always with at least

one large central facet. Sensoria of alate forms linear. Wings with simple
or once-branched cubitus Schizoneurina

Subfamily CHEEMISINAE

1. Eudimentary gonapophyses appearing as three short cones. Wax glands
almost always present Chennisina

— Eudimentary gonapophyses seemingly lacking. Wax glands mostly absent.

Phylloxerina

Group SlPHONOPHORINA

1. Apterous forms with a few sensoria on the third antenual segment. Antennal

tubercles usually well formed. Body almost never with lateral tubercles,
in any case these are never formed on the seventh abdominal segment .... 2

— Apterous forms without sensoria on the third antennal segment. Antennal

tubercles often small or absent. Body with lateral tubercles 4

2. Cornicles almost cylindrical, or rarely somewhat swollen on the side, but then

the body is covered with capitate hairs 3

— Cornicles distinctly clavate. Body almost bare, never with capitate hairs.

Rhopalosiphum Koch
Type Amphorophora ampullata Buckton.

3. Body of apterous forms with long capitate hairs. First antennal joint

drawn out, somewhat tooth-shaped on the inner side Myzus Passerini

Type Aphis ribis Linn.

— Body of apterous forms bare or without capitate hairs. First antennal joint

never drawn out, tooth-like Macrosiphum Passerini

Type Aphis millifolii Fabr.

4. Body of apterous forms with capitate hairs. First antennal joint more or

less toothed on inner side Capitophorus n.gn.

Type Phorodon carduinum Walker.

— Body of apterous forms without capitate hairs. First anteunal joint not

toothed 5

5. Body with many long delicate hairs. Cornicles short, somewhat swollen.

Cladobius (Koch.) Pass.
Type Aphis populea Kalt.

— Body bare or almost so 6

6. Cornicles almost as long or longer than cauda 7

— Cornicles much shorter than cauda 16

7. Cornicles always distinctly clavate Siphocoryne Pass.

Type Aphis avenae Fabr.

— Cornicles cylindrical or conical 8

8. Antennal tubercles well formed, very distinctly toothed on the inner side.

Tubercles on the side of the body always absent 9

— Antennal tubercles mostly small or lacking, never distinctly toothed. Body

often with lateral tubercles 10

9. Antennal tubercles very strongly toothed, the first joint being distinctly

toothed on the inner side Phorodon Pass.

Type Aphis humuli Schr.

— Antennal tubercles only slightly toothed, first antennal joint being rounded

or flat on the inner side, never toothed Ovatus n.gn.

Type Ovatus mespili v. d. G.

156 MISCELLANEOUS STUDIES

10. Antennal tubercles well formed, strongly rounded on the inner side.

Typo Aphis cerasi Fabr. Myzoides n.gn.

— Antennal tubercles small or lacking, never drawn out distinctly on inner

side 11

11. Body with small tubercles on the middle of the seventh and eighth abdominal

segments, and often also on the head and prothorax Dentatus n.gn.

Type Aphis sorbi Kalt.

— Body without tubercles on the middle of the seventh and eighth abdominal

segments 12

12. Cubitus of fore wing only once-branched Toxoptera Koch

Type Toxoptera graminum (Bond.).

— Cubitus of fore wing always twice-branched. Body often with lateral tuber-

cles 13

13. Cornicles short, always distinctly conical. Cauda very short, broaa with

rounded- tip, usually approximately the length of the cornicles, or entirely
lacking. Lateral tubercles lacking or only indistinctly formed on the an-
terior abdominal segments 15

— Cornicles long, almost cylindrical. Cauda sickle- or club-shaped, usually dis-

tinctly shorter than cornicles 14

14. Body long without lateral tubercles. Front often with a very distinct tubercle

in the middle Myzaphls n.gn.

Type Aphis rosarum Walker.

— Body more rounded, with lateral tubercles. Front usually flat, never with a

distinct tubercle Aphis Linn.

Type Aphis rumicis Linn.

15. Cauda distinctly separated, almost as long as broad Brachycaudus n.gn.

Type Aphis myosotidis Koch.

(Aphis cardui Linn, belongs in this genus.)

— Cauda lacking or scarcely separated, much shorter than broad ....Acaudus n.gn.

Type Aphis lychnidis Linn.

16. Cornicles distinctly longer than broad. Cauda usually not conical 17

— Cornicles extremely short, scarcely projecting, cylindrical, usually nearly as

long as broad. Cauda always conical with broad base 19

17. Cornicles only a little longer than broad, distinctly conical. Cauda sickle- or

club-shaped Longicaudis n.gn.

•Type Hyalopterus trirhodus (Walker).

— Cornicles cylindrical, at least twice as long as broad 18

18. Body with lateral tubercles on first and seventh abdominal segments. Cauda

small, club-shaped Hyalopterus Koch

Type Aphis pruni Fabr.

— Body without lateral tubercles on first and seventh abdominal segments. Cauda

conical Serniaphis n.gn.

Type Aphis carotae Koch.

19. Body long, without lateral tubercles. Antennae short, at the most about half

the length of the body Brachycolus Buckton

Type Brachycolus stellariae (Hardy).

— Body oval with lateral tubercles on prothorax, first and seventh abdominal

segments. Antennae at least about three-fourths as long as the body.
Type Aphis thalittri Koch. Brachysiphum n.gn.

Group DREPANOSIPHINA
Genus Drepanosiphum Koch. Type Drepanosiphum platanoides Schrank.

A SYNOPSIS OF THE APHIDIDAE

157

Group CALLIPTERINA

1. Antennae six-jointed. Cornicles merely pores. Body always with wax

glands Phyllaphis Koch

Type Phyllaphis fagi (Linn.).

— Antennae seven-jointed, the terminal process at least one-half as long as the

preceding joint. Cornicles always distinctly projecting. Body almost
always without wax glands, these always of similar shape 2

2. Seventh antennal joint distinctly longer than sixth 3

— Seventh antennal joint only as long as, or shorter than sixth 6

3. Cornicles but slightly projecting. Antennae curved as in beetles.

Bradyaphis Mord.
Type Bradyaphis antennata (Kalt.).

— Cornicles distinctly prominent. Antennae straight 4

4. Anal plate only slightly emarginate, never bilobed. Body with tolerably stiff

hairs, these are never capitate. Apterous forms always with sensoria on

third antennal joint Callipterinella n.gn.

Type Callipterus betularius Kalt.

— Anal plate distinctly bilobed. Body bare or with capitate hairs 5

5. Apterous forms without sensoria on third antennal joint. The body always

with capitate hairs Callipterus Koch

Type Callipterus coryli (Goetze).

— Apterous forms with a few sensoria on third antennal joint. Body with dis-

tinct tubercles Tuberculatus Mord.

Type Tuberculatus "betulicolus (Kalt.).

6. Anal plate distinctly bilobed 3

— Anal plate always simple 7

7. Cauda wart-like, distinctly separated from base. Apterous forms without

sensoria on third antennal joint Callipteroides Mord.

Type Callipterus betulae Koch.

— Cauda scarcely visible. Sensoria present on third antennal joint of apterous

forms Symydobius Mord.

Type Symydobius oblongus (Hey den).

8. Seventh antennal joint nearly as long as the sixth. Wings Avith only very

small black spots at the tip of the veins Subcallipterus Mord.

Type Callipterus alni (Fabr.).

— Seventh antennal joint nearly half as long as sixth. Wings black spotted.

Pterocallis Pass.
Type Pterocallis tiliae (Linn.).

Group CHAITOPHORINA

1. Body with long delicate hairs. Antennae seven-jointed. Cornicles well de-

veloped 2

— Body with short thorn-like hairs. Antennae six-jointed, the terminal process

always distinctly longer than the preceding joint. Cornicles only slightly

projecting Sipha Passerini

Type Sipha glyceriae (Kalt.).

2. Tarsi with two " Haftlappchen " [i.e., the empodial hair is spatula-like].

Chaitophorinella n.gn.
Type Chaitophorus testudinatus (Thornton).

— Tarsi without "Haftlappchen" [i.e. the empodial hair is bristle-like].

Chaitophorus Koch
Type Chaitophorus leucomelas Koch.

158 MISCELLANEOUS STUDIES

Group LACHNINA

1. Wings usually with twice-branched cubitus, the radius always straight. Cauda

not at all or only slightly separated Lachnus 111.

Type Lachnus juniperi De Geer.

— Wings with once- or twice-branchel cubitus and with a curved radius ; the mem-

brane usually with dusky markings. Cauda usually slightly separated 2

2. Beak distinctly longer than the body, strongly retractile. Cubitus but once-

branched, the wings only slightly darkened Stomapnis Buckton

Type Stomaphis quercus (Linn.).

— Beak clearly shorter than body and only slightly retractile. Wings beautifully

spotted with dark brown 3

3. Cubitus twice-branched Dryobius Koch

Type Dryobius croaticus Koch.

— Cubitus once-branched Schizodryobius n.gn.

Type Lachnus exsiccator Hart.

Tribe ANOECIINA

1. Hind tarsi elongate Trama Heyden

Type Trama radicis Koch.

— Hind tarsi not elongate 2

2. Cubitus once-branched. Cornicles present, quite prominent. Margin of body

with peculiar non-faceted "wax-gland" (?) plates Anoecia Koch

Type Anoecia corni (Fabr.).

— Cubitus not branched. Cornicles absent. Wax gland plates not present.

Tullgrenia v. d. G.
Type Tullgrenia phaseoli (Pass.).

Tribe HOEMAPHIDINA

1. Antennae always five-segmented. The fronds almost without exception with

two little horns. Cubitus once branched Cerataphis Licht.

Type C. lataniae Boisd.

— Antennae of the apterae often only three-segmented. Fronds without protuber-

ances. Cubitus simple Hamamelistes Schim.

Type H. betulae Mordw.

A SYNOPSIS OF THE APHIDIDAE 159

APPENDIX 2

HOST PLANT LIST OF CALIFORNIA APHIDIDAE35

Abies (fir)

45. Lachnus ferrisi Swain
47. Laclinus occidentalis Dvdn.
176. Mindarus abietinus Koch
Abutilon (Indian mallow)

146. Aphis senecio Swain
104. Rhopalosiphum persicae (Sulz.)
Acer (maple, box elder)

5. Drepanaphis acerifolii (Thomas)
4. Drepanosiphum platanoides (Schrank)
33. Thomasia negundinis (Thomas)
Achittea (yarrow)

79. Macrosiphum solanifolii (Ashmead)
Aegopodium (goutweed)

108. Siphocoryne capreae (Fabr.)
Aesculus (California buckeye)

88. Mysus circumflexus (Buckton)
171. Prociphilus venafuscus Patch
Alder, see Alnus
Alfalfa, see Medicago
Alfilerilla, see Erodium
Alisma (water plantain)

156. Siphocoryne nymphaeae (Linn.)
Almond, see Prunus
Alnus (Alder)

9. Eucallipterus flava (Davidson)
8. Euceraphis gillettei Davidson
11. Myzocallis alnifoliae (Fitch)
Alopecurus (foxtail)

68. Macrosiphum granarium (Kirby)
88. Myzus circumflexus (Buekton)
Althaea (hollyhock)

121. Aphis euonomi Fabr.
Alum root, see Heuchera

35 In the following list only the generic and common names of the plants are
employed, the various species of plants being omitted. Although in certain cases
aphids are restricted to certain species, as Eriosoma languinosa Hartig on Pyrus
communis but not en Pyrus malus, these are in the minority. The botanical
names are taken from the following works, with preference as in the order listed :

Bailey, L. H., Standard cyclopedia of horticulture, vols. 1-6, New York, Mac-
millan, 1914-1917.

Bobinson, B. L., and Fernald, M. L., Gray's New manual of botany, ed. 7,
Cambridge, Harvard University, 1908.

Jepson, W. L., A flora of western middle California, ed. 2, San Francisco, Cun-
ningham, 1911.

Abrams, LeEoy, Flora of Los Angeles and vicinity, Palo Alto, Stanford Uni-
versity Press, 1904.

160 MISCELLANEOUS STUDIES

Amaranthus (pigweed)

123. Aphis gossypii Glover

132. Aphis middletonii Thomas

104. Bhopalosiphum persicae (Sulz)

163. Trifidaphis radicicola (Essig)
Ambrosia (ragweed)

146. Aphis senecio Swain

77. Macrosiphum rudbeckiae (Fitch)
Ampelodesma

68. Macrosiphum granarium (Kirby)
Amsinckia (amsinckia)

146. Aphis senecio Swain

104. Bhopalosiphum persicae (Sulz.)
Angelica (angelica)

109. Aphis angelicae Koch

115. Aphis cari Essig

108. Siphocoryne capreae (Fabr.)
Anise, Wild, see Carum
Anthemis (chamomile)

121. Aphis euonomi Fabr.
146. Aphis senecio Swain
123. Aphis gossypii Glover

Apple, see Pyrus
Apricot, see Prunus
Aquilegia (columbine)

85. Myzus aquilegiae Essig
Arbor vitae, see Thuja sp.
Arbutus (madrone, strawberry tree)

103. Rhopalosiphum nervatum Gillette
Arctostaphylos (manzanita)

1. Phyllaphis coweni (Cockerell)

103. Bhopalosiphum nervatum Gillette
Artemisia (sagebrush, oldman, California mugwort, etc.)

122. Aphis frigidae Oestlund
131. Aphis medicaginis Koch
137. Aphis oregonensis Wilson
146. Aphis senecio Swain

61. Macrosiphum artemisiae (Fonsc.)

62. Macrosiphum artemisioola (Williams)
72. Macrosiphum ludovicianiae (Oestlund)

Artichoke, see Cynara
Arundinaria (bamboo)

12. Myzocallis arundicolens (Clarke)

13. Mysocallis arundinariae Essig
Arundo (giant reed)

12. Mysocallis arundicolens (Clarke)

13. Mysocallis arundinariae Essig
Asclepias (milkweed)

123. Aphte gossypii Glover
135. Aphis nerii Fonsc.

102. Bhopalosiphum lactucae (Kalt.)

A SYNOPSIS OF THE APH1DIDAE 161

Ash, see Fraxinus

Arparagus (asparagus, smilax, asparagus fern)

123. Aphis gossypii Glover

88. Myzus circumflexus (Buckton)
Aster (aster)

132. Aphis middletonii Thomas

146. Aphis senccio Swain
Astragalus (loco weed)

131. Aphis medicaginis Koch
Atriplex (oraehe)

147. Aphis tetrapteralis Cockerell

79. Macrosiphum solanifolii (Ashmead)
Avocado, see Persea
Avena (oats)

111. Aphis avenae Fabr.

68. Macrosiphum granarium (Kirby)
105. Rhopalosiphum rhois Monell
Baccharis ( groundsel )

146. Aphis senecio Swain

63. Macrosiphum baccharadis (Clarke)

77. Macrosiphum rudbeckiae (Fitch)
104. Rhopalosiphum persicae (Sulz.)
Bamboo, see Arundinaria, Bambusa, and Phyllostachys
Bambusa (bamboo)

12. Myzocallls arundicolens (Clarke)
Banana, see Musa
Barberry, see Berberis
Barley, see Hordeum
Basswood, see Tilia
Bean, see Pliaseolus
Bean, Blackeye, see Vigna
Bean, Broad, see Vicia
Beech, see Fagus
Beet, see Beta
Begonia (begonia)

123. Aphis gossypii Glover
Bell, fairy, see Dipsorum
Berberis (barberry)

154. Liosomaphis berberidis (Kalt.)
Beta (beet, sugarbeet)

124. Aphis gossypii Glover
164. Pemphigus betae Doane

Betula (birch)

6. Calaphis betulaecolens (Fitch)
27. Callipterinella annulata (Koch)

7. Euceraphis betulae (Koch)
Birch, see Betula

Blackberry, see Rubus
Bougainvillea (bougainvillea)

104. Rhopalosiphum persicae (Sulz.)
Boxelder, see Acer

162 MISCELLANEOUS STUDIES

Brassica (cabbage, mustard, turnip, etc.)

112. Aphis brassicae Linn.

144. Aphis pseudobrassicae Davis

166. Pemphigus populi-transversus Riley (?)

102. Rhopalosiphum lactucae (Kalt.)

104. Rhopalosiphum persicae (Sulz.)

Broom, see Cytisus

Buckeye, California, see Aesculus

Buckton, see Rhamnus

Bur clover, see Medicago

Buttercup, see Ranunculus

Cabbage, see Brassica

Calendula (marigold)

113. Aphis calendulicola Monell
121. Aphis euonomi Fabr.
146. Aphis senecio Swain

California buckeye, see Aesculus
California holly, see Heteromeles
California mugwort, see Artemesia
California poppy, see Esclischoltsia
California sagebrush, see Artemisia
California tule, see Typlia
Calla, see Zantcdesohia
Camellia (camellia)

152. Toxoptera aurantii (Fonsc.)
Canary grass, see Phalaris
Cantaloupe, see Cucumis
Capsella (shepard's purse)

112. Aphis brassicae Linn.

123. Aphis gossypii Glover

141. Aphis pseudobrassicae Davis

104. Rhopalosiphum persicae (Sulz.)
Capsicum (pepper pimento)

104. Rhopalosiphum persicae (Sulz.)
Caragana (pea tree)

131. Aphis medicaginis Koch.
Carum (wild anise)

115. Aphis cari Essig

155. Siphocoryne capreae (Fabr.)
Castanea (chestnut)

15. Myzocallis castanicola Baker (davidsoni Swain)
Catalpa (catalpa)

123. Aphis gossypii Glover

139. Aphis pomi de Geer

104. Rhopalosiphum persicae (Sulz.)
Cauliflower, see Brassica
Ceanothus (mountain lilac)

116. Aphis ceanothi Clarke
Centaurea (tacalote)

130. Aphis marutae Oestlund
Centranthus (red valerian)

76. Macrosiphum rosae (Linn.)
104. Rhopalosiphum persicae (Sulz.)

A SYNOPSIS OF THE APHIDIDAE 163

Chaerophyllum

155. Siphocoryne capreae (Fabr.)
Chamomile, see Anthemis
Chaparral broom, see Baccharis
Charlock, see Brassica
Cheeseweed, see Malva
Cheiranthus (wallflower)

88. Myzus circumflexus (Buckton)
Chenopodium (lamb's-quarters, pigweed)

110. Aphis atriplicis Linn.

123. Aphis gossypii Glover

124. Aphis hederae Kalt. ( ?)
164. Pemphigus betae Doane (?)
104. Rhopalosiphum persicae (Sulz.)

Cherry, see Prunus
Cherry, wild, see Prunus
Chestnut, see Castanea
Chestnut, Horse, see Aesculus
Chicory, see Chicorium
Christmas berry, see Heteromeles
Chrysanthemum (chrysanthemum)

56. Amphorophora latysiphon Davidson

123. Aphis gossypii Glover

146. Aphis senecio Swain

160. Coloradoa rufomaculata Wilson

77. Macrosiphum rudbeckiae (Fitch)

78. Macrosiphum sariborni Gillette
Cichorium (chicory)

71. Macrosiphum lactucae (Kalt.)
Cicuta (water hemlock)

155. Siphocoryne capreae (Fabr.)
Cirsium (thistle)

144. Aphis cardui Linn.
Citrullus (watermelon)

123. Aphis gossypii Glover
Citrus (citrus, orange, lemon, etc.)

118. Aphis cooM Essig

123. Aphis gossypii Glover

131. Aphis medicaginis Koch

79. Macrosiphum solanifolii (Ashmead)
104. Rhopalosiphum persicae (Sulz.)
152. Toxoptera aurantii (Fonsc.)

Clarkia (clarkia)

104. Rhopalosiphum persicae (Sulz.)
Clematis (clematis)

94. Myzus varians Davidson
Clover, see Trifolium
Clover, Sweet, see Melilotus
Coffeeberry, see Rhamnus
Columbine, see Aquilegia
Compositae (various species)

131. Aphis medicaginis Koch

65. Macrosiphum chrysanthemi (Oestlund)

77. Macrosiphum rudbeckiae (Fitch)

164 MISCELLANEOUS STUDIES

Conium (poison hemlock)

159. Siplwcoryne pastinacae (Linn.)
Convolvulus (morning glory)

56. Amphorophora latysiplwn Davidson
123. Aphis gossypii Glover
72. Macrosiphum ludovicianae (Oestlund)
Corn, see Zea
Cornus (dogwood)

119. Aphis cornifoliae Fitch
123. Aphis gossypii Glover

Corylus (hazelnut)

16. Mysocallis coryli (Goetze)

100. Bhopalosiphum corylinum Davidson
Cotoneaster (cotoneaster)

139. Aphis pomi de Geer
Cotton, see Gossypium
Cottonwood, see Populus
Cow parsnip, see Heracleum
Cowpea, see Vigna
Cowslip, see Primula
Crab apple, see Pyrus
Cranesbill, see Geranium
Crataegus (hawthorn)

120. Aphis crataegifoliae Fitch
139. Aphis pomi de Geer

Cruciferae (various spp.)

112. Aphis brassicae Linn.

141. Aphis pseudobrassicae Davis
Cucumber, see Cucumis
Cucumis (cucumber, muskmelon, cantaloupe, etc.)

123. Aphis gossypii Glover
Cucurbita (squash, gourd, pumpkin, etc.)

123. Aphis gossypii Glover

66. Macrosiphum cucurbitae (Thomas)
Cupressus (cypress)

161. Cerosipha cupressi Swain

96. Macrosiphum morrisoni Swain
Currant, see Eibes
Cydonia (quince)

139. Aphis pomi De Geer
Cynara (artichoke)

86. Myzus braggii Gillette
Cynoglossum ( houndstongue )

104. Elwpalosiphum persicae (Sulz.)
Cypress, see Cupressus
Cyrtomium (holly fern)

162. Cerataphis lataniae (Boisd.)
58. Idiopterus nephrelepidis Davis
88. Myzus circumflexus (Buckton)

Cytisus (broom)

146. Aphis senecio Swain

104. Ehopalosiphum persicae (Sulz.)

A SYNOPSIS OF THE APHIDIDAE 165

Dandelion, see Taraxacum
Datura (jimson weed)

123. Aphi# gossypii Glover
Deinandra

79. Macrosiplmm solanifolii (Ashmead)
Digitalis (foxglove)

88. Myzus circumflexus (Buckton)
Dipsacus (fuller's teasel)

7G. Macrosiphum rosae (Linn.)

77. Macrosiphum rudbeckiae (Fitch)

104. Ehopalosiphum persicae (Sulz.)
Disporum (fairy bell)

79. Macrosiphum solanifolii (Ashmead)
Dock, see Rumex
Dogwood, see Cornus
Douglas fir, see Pseudotsuga
Dracaena (dragon tree)

111. Aphis avenae Fabr.
Dragon tree, see Dracaena
Eldei berry, see Sambucus
Elm, see Ulmus
Elymus (wild rye)

68. Macrosiphum granarium (Kirby)
English ivy, see Hedera
Epilobium (fireweed)

136. Aphis oenotherae Oestlund
Eriobotrya (loquat)

139. Aphis pomi de Geer
Erodium (alfilerilla)

79. Macrosiphum solanifolii (Ashmead)

104. Ehopalosiphum persicae (Sulz.)
Erysimum (western wallflower)

155. Siphocoryne capreae (Fabr.)
Escallonia (escallonia)

104. Ehopalosiphum persicae (Sulz.)
Eschscholtzia (California poppy)

123. Aphis gossypii Glover
Everlasting, see Gnaphalium
Fagus (beech)

2. Phyllaphis fagi (Linn.)
Fairybell, see Dipsorum
Fennel, see Foeniculum
Fenugreek, see Trigonella
Fern, asparagus, see Asparagus
Fern, Boston, see Nephrolepis
Fern, holly, see Cyrtomium
Fig marigold, see Mesembryanthemum
Figwort, see Scrophularia
Fir, see Abies

Fir, Douglas, see Pseudotsuga
Fireweed, see Epilobium
Foeniculum (fennel)

155. Siphocoryne capreae (Fabr.)

166 MISCELLANEOUS STUDIES

Foxglove, see Digitalis

Foxtail, see Alopecurus

Fragaria (strawberry)

90. Myzus fragaefolii Cockerell

Fraxinus (ash)

171. Prociphilus venafuscus Patch

167. Thecabius californicus (Davidson)

Fuller's teasel, see Dipsacus

Fuchsia (fuchsia)

79. Macrosiphum solanifolii (Ashmead)
88. Myzus circumflexus (Buckton)

Gambleweed, see Sanicula

Geranium, see Pelargonium

Geranium (cranesbill)

104. Rhopalosiphum persicae (Sulz.)
German ivy, see Senecio

Gladiolus (gladiolus)

88. Myzus circumflexus (Buckton)
Glycyrrhiza (liquorice)

131. Aphis medicaginis Koch
Gnaphalium (everlasting)

146. Aphis senecio Swain

60. Macrosiphum ambrosiae (Thomas)
Gooseberry, see Ribes
Goosefoot, see Chenopodium
Gossypium (cotton)

123. Aphis gossypii Glover
Gourd, see Cucurbita
Goutweed, see Aegopodium
Graminaceae (various species)

111. Aphis avenae Fabr.

67. Macrosiphum dirhodum (Walker)

68. Macrosiphum granarium (Kirby)

105. Rhopalosiphum rhois Monell
Grape, see Vitis

Grindelia (marsh grindelia)

146. Aphis senecio Swain
Hawthorn, see Crataegus
Hazelnut, see Corylus
Hcdera (English ivy)

109. Aphis angelicae Koch

124. Aphis hederae Kalt

104. Rhopalosiphum persicae (Sulz.)
Hedge mustard, see Sisymbrium
Hedge nettle, see Stachys
Helianthus (sunflower)

123. Aphis gossypii Glover

132. Aphis middletonii Thomas
146. Aphis senecio Swain

60. Macrosiphum ambrosiae (Thomas)
77. Macrosiphum rudbeckiae (Fitch)
104. Rhopalosiphum persicae (Sulz.)

A SYNOPSIS OF THE APHIDIDAE 167

Hemlock, Poison, see Conium
Hemlock, Water, see Cicuta
Heracleum (cow parsnip)

123. Aphis gossypii Glover

125. Aphis heraclei Cowen
Heteromeles (California holly, Christmas berry)

170. Prociphilus alnifoliae (Williams)

103. Ehopalosiphum nervatum Gillette
Heuchera (alum root)

69. Macrosiphum heucherae (Thomas)
Hibiscus (rose mallow)

121. Aphis euonomi Fabr.
Holly fern, see Cyrtonium
Hollyhock, see Althaea
Holly, mountain, see Heteromeles
Honey flower, see Melianthus
Honeysuckle, see Lonicera
Hop, see Humulus
Hordeum (barley)

111. Aphis avenae Fabr.

68. Macrosiphum granarium (Kirby)
Houndstongue, see Cynoglossum
Humulus (hop)

123. Aphis gossypii Glover
98. Phorodon humuli (Schrank)
Hydrangea (hydrangea)

123. Aphis gossypii Glover
Indian mallow, see Abutilon
Ironweed, see Veronina
Ivy, Engislh, see Hedera
Ivy, German, see Senecio
Jasminum (jessamine)

70. Macrosiphum jasmini (Clarke)
Jessamine, see Jasminum

Jimpson weed, see Datura
Juglans (walnut)

24. Callipterus calif ornicus (Essig)

25. Callipterus caryae Monell

23. Chromaphis juglandicola (Kalt.)

26. Monellia caryella (Fitch)
Knotweed, see Polygonum
Lactuca (lettuce)

79. Macrosiphum solanifolii (Ashmead)
Lamb 's-quarters, see Chenopodium
Lathyrus (sweet pea)

74. Macrosiphum pisi (Kalt.)
Laurel, see Laurus
Laurel, California, see Umbellularia
Laurestinus, see Viburnum
Laurus (laurel)

150. Aphis viburnicolens n.sp.

168 MISCELLANEOUS STUDIES

Lavatcra (tree mallow)

104. Khopalosiphum persicae (Sulz.)
Leather root, see Psorales
Lemon, see Citrus
Lcpidium (peppergrass)

123. Aphis gossypii Glover
Lettuce, see Lactuca
Ligusticum (lovage)

155. Siphocorync capreae (Fabr.)
Lilac, see Syringa
Lilac, Mountain, see Ceanothus
Lilium (lily)

123. Aphis gossypii Glover
88. Myzus circumflexus (Buekton)
Lily, see Lilium
Lily, Water, see Nymphaea
Linden, see Tilia
Liquorice, see Glycyrrhiza
Liriodendron (tulip tree)

104. Bhopalosiphum persicae (Sulz.)
Lithospermum

127. Aphis lithospermi Wilson
Loco weed, see Astragalus
Locust, see Bobinia
Loganberry, see Bubus
Lonicera (honeysuckle)

157. Siphocoryne pastinacae (Linn.)
Loquat, see Eriobotrya
Lovage, see Ligusticum
Lupinus (lupine)

59. Macrosiphum albifrons Essig
Lycopersicum (tomato)

91. Myzus lycopersicae (Clarke)

104. Bhopalosiphum persicae (Sulz.)
Madia (tarweed)

77a. Macrosiphum rudbeclciae '(Fitch) var. madia n.var.
Madron, see Arbutus
Mallow, Indian, see Abutilon
Mallow, Rose, see Hibiscus
Mallow tree, see Lavatera
Malva (cheeseweed)

121. Aphis euonomi Fabr.

123. Aphis gossypii Glover

104. Bhopalosiphum persicae (Sulz.)
Manzanita, see Arctostaphylos
Maple, see Acer
Marigold, see Calendula,
Marigold, fig, see Mescmbryanthemum
Matthiola (ten-weeks' stock)

141. Aphis pseudobrassicae Davis
Mayten, see Maytenus
Maytenus (mayten)

121. Aphis euonomi Fabr.

A SYNOPSIS OF THE APHIDIDAE 169

Medicago (alfalfa, bur clover, etc.)

131. Aphis medicaginis Koch

74. Macrosiphum pisi (Kalt.)
Melianthus (honey flower)

104. Bhopalosiphum persicae (Sulz.)
Melilotus (sweet clover)

131. Aphis medicaginis Koch
Mesembryanthemum (fig marigold)

121. Aphis euonomi Fabr. •
Milk thistle, see Silybum
Milkweed, see Asclepias
Morning glory, see Convolvulus
Morus (mulberry)

133. Aphis mori Clarke
Mountain holly, see Heteromeles
Mountain lilac, see Ceanothus
Mugwort, California, see Artemisia
Mulberry, see Morus
Musa (banana)

111. Aphis avenae Fabr.
Muskmelon, see Cucumis
Mustard, see Brassica
Mustard, Hedge, see Sisymbrium
Mustard, Teasel, see Erysimum
Nasturtium, see Tropaeolum
Nectarine, see Prunus
Nephrolcpis (Boston fern)

58. Idiopterus nephrelepidis Davis
Nerium (oleander)

135. Aphis nerii Fonsc.
Nettle, Hedge, see Stachys
Nettle, Stinging, see Urtica
Nightshade, see Solanum
Ninebark, see Physocarpus
Nymphaea (water lily)

123. Aphis gossypii Glover

156. Siplwcoryne nymphaeae (Linn.)
Oak, see Quercus
Oak, Poison, see Ehus
Oak, Tanbark, see Pasania
Oats, see Avena
Oenothera (evening primrose)

136. Aphis oenotherae Oestlund
Oldman, see Artemisia

Oleander, see Nerium
Orange, see Citrus
Orache, see Atriplex
Orchidaceae (orchids)

162. CeratapJns lataniae (Boisd.)
Orthocarpus (owl clover)

73. Macrosiphum orthocarpi Davidson
Owl clover, see Orthocarpus

170 MISCELLANEOUS STUDIES

Oxalis (oxalis)

97. Macrosiphum solanifolii (Ashmead)
104. Rhopalosiphum persicae (Sulz.)

152. Toxoptera aurantii (Fonsc.)
Pansy, see Viola

Papaver (poppy)

121. Aphis euonomi Fabr. (papaveris Fabr.?)
Parsley, see Petroselinum
Parsnip, see Pastinaca
Parsnip, Cow, see Heracleum
Pasania (tanbark oak)

20. Myzocallis pasaniae Davidson
Pastinaca (parsnip)

156. Siphocoryne pastinacae (Linn.)
Pea, see Pisum
Pea, Cow, see Vigna
Pea, Sweet, see Lathyrus
Pea tree, see Caragona
Peach, see Prunus
Pear, see Pyrus
Pelargonium (geranium)

97. Pentalonia nigronervosa Coquellet

104. Rhopalosiphum persicae (Sulz.)
Pentstemon (pentstemon)

88. Myzus circumftexus (Buckton)

104. Rhopalosiphum persicae (Sulz.)
Pepper, see Capsicum
Peppergrass, see Lepidium
Periwinkle, see Vinca
Per sea (avacado)

123. Aphis gossypii Glover
Petroselinum (parsley)

155. Siplwcoryne capreae (Fabr.)
Phalaris (canary grass)

111. Aphis avenae Fabr.

153. Hyalopterus arundinis (Fabr.)
Phaseolus (bean)

121. Aphis euonomi Fabr. (rumicis Linn.?)
131. Aphis medicaginis Koch

74. Macrosiphum pm (Kalt.)
Phragmites (reed grass)

153. Hyalopterus arundinis (Fabr.)
Phyllostachys (bamboo)

12. Myzocallis arundicolens (Clarke)
Physocarpus (ninebark)

100. Rhopalosiphum corylinum Davidson
Pice a (spruce)

46. Lachnus glehnus Essig
55. Lachnus vanduzei n.sp.
158. Myzapliis abietina (Walker)
Pigweed, see Amaranthus, and Chenopodium
Pimento, see Capsiwm

A SYNOPSIS OF THE APHIDIDAE 171

Pimpinella

155. Siphocoryne capreae (Fabr.)
Pine, see Pinus

Pinus (pine)

178. Chermes cooleyi Gillette

179. Chermes pinicorticis Fitch
43. Essigella californica (Essig)
45. Laclmus ferrisi Swain

48. Lachnus oregonensis Wilson

49. Lachnus pini-radiatae Davidson

50. Lachnus .ponderosa Williams
52. Lachnus sabinianus n.sp.

53a. Lachnus tomentosa (De Geer) (Addenda)
176. Mindarus abietinus Koch
Pisum (pea)

74. Macrosiphum pisi (Kalt.)
Pittosporum (pittosporum)

139. Aphis pomi De Geer

79. Macrosiphum solanifolii (Ashmead)
104. Jlhopalosii>hum persicae (Sulz.)
Plantago (plantain)

123. Aphis gossypii Glover
129. Aphis malifoliae Fitch (?)

88. Hyzus circumflexus (Buckton)
Plantain, see Plantago
Plantain, Water, see Alisma
Plat anus (western sycamore)

4. Drepanosiphum platanoides (Schrank)
Plum, see Prunus
Polygonum (knotweed)

100. Rhopalosiphum hippophaes (Koch).

156. Siphocoryne nymphaeae (Linn.)
Pomegranate, see Punica

Pondweed, see Potamogeton

Poplar, see Populus

Poppy, see Papaver

Poppy, California, see Eschscholtzia

Populus (poplar, cottonwood)

28. Arctaphis populifolii (Essig)

164. Pemphigus bctae Doane

165. Pemphigus populi-caulis Fitch

166. Pemphigus populi-transversus Eiley
181. Phylloxerina popularia (Pergande)

40. Ptcrocomma populifoliae (Fitch)

167. Thecabius calif ornicus (Davidson)

168. Thecabius populiconduplifolius (Cowen)

169. Thecabius populi-monilis (Eiley)

34. Thomasia populicola (Thomas)

35. Thomasia salicola (Essig)
Potamogeton (pondweed)

156. Siphocoryne nymphaeae (Linn.)
Potato, see Solarum

172 MISCELLANEOUS STUDIES

Primrose, Evening, see Oenothera
Primula (cowslip)

56. Amphorophora latysiphon Davidson
Prune, see Prunus
Prunus (almond, apricot, cherry, nectarine, peach, plum, prune)

107. Aphis alamedensis Clarke

114. Aphis cardui Linn.

117. Aphis cerasifoliae Fitch

138. Aphis persicae-niger Smith
140. Aphis prunorum Dobr.

153. Eyalopterus arundinis (Fabr.

87. Myzus cerasi (Fabr.)

98. Phorodon humuli (Schrank)
104. Bhopalosiphum persicae (Sulz.)
156. Siphocoryne nymphaeae (Linn.)
Pseudotsuga (Douglas fir)

178. Chermes cooleyi Gillette

43. Essigella californica (Essig)

51. Lachnus pseudotsuga Wilson

53. Lachnus taxifolia Swain
171. Prociphilus venafuscus Patch
Psorales (leather root)

74. Macrosiphum pisi (Kalt.)
Pteris (brake)

75. Macrosiphum pteridis Wilson
Punica (pomegranate)

123. Aphis gossypii Glover
Pumpkin, see Cucur~blta
Pyrus (apple, pear)

123. Aphis gossypii Glover

129. Aphis malifoliae Fitch

139. Aphis pomi De Geer

175. Eriosoma languinosa Hartig (pyricola B. & D.)
174. Eriosoma lanigerum (Hausman)
Quercus (oak)

5. Drepanaphis acerifolii (Fitch) (T)

14. Myzocallis bellus (Walsh)

15. Myzocallis castanicola Baker (davidsoni Swain)

17. Myzocallis discolor (Monell)

18. Myzocallis punctatus (Monell)

19. Myzocallis californicus Baker (maureri Swain)
21. Myzocallis quercus (Kalt.)

3. Phyllaphis quercicola Baker

36. Symydobius agrifoliae Essig

37. Symydobius chrysolepis Swain
177. Vacuna dryophila Schrank (f)

Quince, see Cydonia
Eadish, see Eaphanus
Eagweed, see Ambrosia
Ramona (black sage)

142. Aphis ramona Swain

A SYNOPSIS OF THE APHIDIDAE 173

Ranunculus (buttercup)

132. Aphis middletonii Thomas
104. Rhopalosiphum persicae (Sulz.)

167. Thecabius californicus (Davidson)

168. Thecabius populiconduplifolius (Cowen)
Raphanus (radish)

112. Aphis brassicae Linn.

141. Aphis pseudobrassicae Davis

104. Rhopalosiphum persicae (Sulz.)
Eeed, Giant, see Arundo

Eeed grass, see Phragmites
Rhamnus (buckthorn, coffeeberry)
123. Aphis gossypii Glover

92. Myzus rhamnus (Clarke)
Rhus (poison oak)

105. Rhopalosiphum rhois Monell
Ribes (currant, gooseberry)

126. Aphis houghtonensis Troop
134. Aphis neo-mexicana Ckll. var. pacifica Dvdn.
89. Myzus cynosbati (Oestlund)

93. Myzus ribifolii Davidson
Robinia (locust)

131. Aphis medicaginis Koch
Rosa (rose, wild and cultivated)

67. Macrosiphum dirhodum (Walker)

76. Macrosiphum rosae (Linn.)
159. Myzaphis rosarum (Walker)

103. Myzus nervatum Gillette
Eose, see Rosa

Rose mallow, see Hibiscus

Rubus (blackberry, loganberry, thimbleberry)

143. Aphis rubiphila Patch

57. Amphorophora rubi (Kalt.)

95. Nectarosiphon rubicola (Oestlund)
Rumex (dock, sorrell)

121. Aphis euonomi Fabr. (rumicis Linn.)

123. Aphis gossypii Glover

146. Aphis senecio Swain

164. Pemphigus betae Doane (?)

104. Rhopalosiphum persicae (Sulz.)
Eye, Wild, see Elymus

Sagebrush, see Artemisia
Sage, Black, see Ramona
Salix (willow)

144. Aphis salicicola Thomas
146. Aphis senecio Swain

29. Arctaphis viminalis (Monell)
31. Fullaicaya saliciradicis Essig

64. Macrosiphum calif ornicum (Clarke)

30. Micrella monella Essig

182. Phylloxerina salicola (Pergande)
40. Pterocomma flocculosa (Weed)

174 MISCELLANEOUS STUDIES

41. Ptcrocomma populifoliae (Fitch)

42. Pterocomma smithiae (Monell)
155. Siphocoryne capreae (Fabr.)

38. Symydobius macrostachyae Essig

39. Symydobius salicicorticis Essig
32. Thomasia crucis Essig

34. Thomasia populicola (Thomas)

35. Thomasia salicola (Essig)

44. Tuberolachnus viminalis (Fonsc.)
Sambucus (elderberry)

145. Aphis sambucifoliae Fitch

81. Macrosiphum stanleyi Wilson
104. Ehopalosiphum persicae (Sulz.)
Sanicula (gambleweed)

119. Aphis cornifoliae Fitch
104. Ehopalosiplium persicae (Sulz.)
Scrophularia (figwort)

99. Phorodon scrophulariae Thames
Senecio (German ivy, ivy senecio)
144. Aphis senecio Swain

88. Myzus circumflexus (Buckton)
104. Ehopalosiphum persicae (Sulz.)
Shepherd 's-purse, see Capsella
Silybum (milk thistle)

121. Aphis euonomi Fabr.
130. Aphis marutae Oestlund
Sisymbrium (hedge mustard)

88. Myzus circumflexus (Buckton)
Smilax, see Asparagus
Snowball, see Virburnum
Snowberry, see Symphoricarpos
Solatium (potato, nightshade)

56. Amphorophora laty siphon Davidson
79. Macrosiplium solanifolii (Ashmead)
88. Myzus circumflexus (Buckton)
102. Ehopalosiphum lactucae (Kalt.)
104. Ehopalosiphum persicae (Sulz.)
163. Trifldaphis radicicola (Essig)
Sonchus (sow thistle)

79. Macrosiphum solanifolii (Ashmead)

80. Macrosiplium sonchella (Monell)
102. Ehopalosiphum lactucae (Kalt.)
104. Ehopalosiphum persicae (Sulz.)

Sorghum

129. Aphis maidis Fitch
Sorrell, see Eumex
Sow thistle, see Sonchus
Spinacia (spinach)

123. Aphis gossypii Glover

104. Ehopalosiphum persicae (Sulz.)
Spirea (spirea)

148. Aphis spiraecola Patch

A SYNOPSIS OF THE APH1DIDAE 175

Spruce, see Picea
Squash, see Cucurbita
Stachys (hedge nettle)

73. Macrosiphum ludovicianae (Oestlund)
88. Myzus circumflexus (Buckton)
Stock, Ten-week, see Matthiola
Strawberry, see Fragaria
Strawberry tree, see Arbutus
Sugar beet, see Beta
Sunflower, see Helianthus
Sweet clover, see Melilotus
Sweet pea, see Latliyrus
Sycamore, Western, see Platanus
Symplwricarpos (snowberry)

108. Aphis albipes Oestlund
Syringa (lilac)

131. Aphis medicaginis Koch
Tacalote, see Centaurea
Taraxacum (dandelion)

82. Macrosiplium taraxici (Kalt.)
Tarweed, see Madia and Hemizonia
Teasel, Fuller's, see Dipsacus

Teasel, mustard, see Erysimum
Thimbleberry, see Eubus
Thistle, see Cirsium
Thistle, Milk, see Silybum
Thistle, Sow, see Sonchus
Thuja (arbor vitae)

54. Lachnus tujafilinus (Del Guercio)
Tilia (linden, basswood)

10. Eucallipterus tiliae (Linn.)
Tomato, see Lycopersicum
Tri folium (clover)

146a. Aphis bakeri Cowen
131. Aphis medicaginis Koch
Trigonella (fenugreek)

69. Macrosiphum pisi (Kalt.)
Triticum (wheat)

111. Aphis avenae Fabr.

64. Macrosiphum granarium (Kirby)
Tropaeolum (nasturtium)

121. Aphis euonomi Fabr.

88. Myzus circumflexus (Buckton)
104. Ehopalosiphum pcrsicae (Sulz.)
Tule, California, see Typha
Tulip, see Tulipa
Tulip tree, see Liriodendron
Tulipa (tulip)

83. Macrosiphum tulipae (Monell)
104. Ehopalosiphum pcrsicae (Sulz.)

Turnip, see Brassica

176 MISCELLANEOUS STUDIES

Typha (California tule)

111. Aphis avenae Fabr.
153. Hyalopterus arundinis (Fabr.)
68. Macrosiphum granarium (Kirby)

156. Siphocoryne nymphaeae (Linn.)
Ulmus (elm)

28. Arctaphis populifolii (Essig) (?)

172. .Colopha ulmicola (Fitch)

173. Eriosoma americana (Eiley)

175. Eriosoma languinosa Hartig (pyricola B. & D.)

174. Eriosoma lanigerum (Hausman)

79. Macrosiphum solanifolii (Ashmead)
22. Myzocallis ulmifolii (Monell)
Umbellularia (California laurel)

88. Myzus circumflexus (Buckton)
104. Ehopalosiphum pcrsicae (Sulz.)

157. Siphocoryne pastinacae (Linn.)
Urtica (stinging nettle)

121. Aphis euonomi Fabr.
Valerian, Bed, see Centranthus
Valeriana

84. Macrosiphum valerianae (Clarke)
Vernonia (ironweed)

123. Aphis gossypii Glover
Vetch, see Vicia

Viburnum (lauristinus, snowball)
121. Aphis euonomi Fabr.
139. Aphis pomi De Geer
150. Aphis viburnicolens n.sp.
Vicia (horse bean, vetch)

121. Aphis euonomi Fabr. (fabae Scop.)
131. Aphis medicaginis Koch

74. Macrosiphum pisi (Kalt.)
Vigna (blackeye bean, cowpea)

131. Aphis medicaginis Koch
Vineca (periwinkle)

56. Amphorophora latysiphon Davidson
88. Myzus circumflexus (Buckton)
104. Ehopalosiphum persicae (Sulz.)
Viola (pansy, violet)

58. Idiopterus nephrelepidis Davis
74. Macrosiphum pisi (Kalt.)
88. Myzus circumflexus (Buckton)
106. Ehopalosiphum violae Pergande
Vitis (grape)

180. Phylloxera vitifoliae (Fitch)
Wallflower, ;,ee Cheiranthus
Wallflower, Western, see Erysimum
Walnut, see Juglans
Water hemlock, see Cicuta
Watermelon, see Citrullus
Water plantain, see Alisma

A SYNOPSIS OF THE APHIDIDAE 177

Wheat, see Triticum
Willow, see Salix
Yarrow, see Achillea
Yucca (yucca)

151. Aphis yuccae Cowen
Zantedeschia (calla)

88. Myzus circumflexus (Buckton)
Zea (corn)

111. Aphis avenae Fabr.

128. Aphis maidis Fitch
Zizia

155. Siphocoryne capreae (Fabr.)

178 MISCELLANEOUS STUDIES

ADDENDA

Since the preparation of this manuscript there have appeared a few papers36
in which there are some new records for certain of the California Aphididae
and in which there are notes concerning the synonymy of some of the species.
These records are noted here and are listed in the Host Plant Index (appendix 2).

2. Phyllaphis fagl (Linn.) on Fagus tricolor, Oakland (Essig, p. 321).

7. Euceraphis betulae (Koch) on Betula populifolia laciniata and B. papy-
rifera (Essig, pp. 322-323).

10. Eucallipterus tiliae (Linn.) on Tilia tomentosa, Berkeley (Essig, p. 323).
Baker places this species in the genus Myzocallis, for although it is quite distinct
from the type of Myzocallis, various species form definite connections leading to
this one.

15. Myzocallis castanicola Baker .(Baker, p. 424). This name has been sug-
gested by Baker to replace M. castaneae (Buckton) (preoccupied by castaneae
(Fitch)). Therefore the name suggested by the author, M. davidsoni Swain,
must be dropped. Essig (p. 323) lists M. castaneae (Fitch), but he refers to this
species.

19. Myzocallis californicus Baker (Baker, pp. 421-422). This is the same
species as described by the author under the name, Myzocallis maureri Swain,
which name will have to be dropped, and replaced by M. californicus Baker.

53a. Lachnus tomentosus (De Geer), on Finns radiata, Berkeley (Gillette,
pp. 140-141). This species is very similar to L. pini-radiatae Davidson, accord-
ing to Gillette. The author finds on looking over his specimens that some of them
labeled L. pini-radiatae Dvdn. are this species, particularly those taken on the
campus at Berkeley.

56. Amphorophora latysiphon Davidson, on Chrysanthemum and Primula sp.,
Berkeley (Essig, p. 329).

68. Macrosiphum granarium (Kirby), on Alopecurus pratensis, Ampelodesma
tenax, and Elymus sp., Martinez (Essig, p. 328).

76. Macrosiphum rosae (Linn.) on Dipsacus fullonum and Centranthus ruber,
Berkeley (Essig, p. 329).

79. Macrosiphum solanifolii (Ashmead), on Achillea millefolium and Pitto-
sporum tobira, Berkeley, and on TJlmus americanus, San Francisco (Essig, p. 329).

88. Myzus circumflexus (Buckton), on Lilium spp., Pentstemon spcctabilis,
and Umbellularia calif ornica, Berkeley (Essig, p. 335).

102. Rhopalosiphum lactucae (Kalt.). Dobrovliansky lists this as a synonym
of E. ribis (Buckton), giving the latter name preference.

so Baker, A. C., Eastern aphids, new and little known, II, Jour. Econ. Ent.,
vol. 10, pp. 421-433, 1917.

Baker, A. C.. The correct name for our apple-grain aphis, Science, vol. 46,
pp. 410-411, 1917.

Davidson, W. M., The reddish-brown plum aphis, Jour. Econ. Ent., vol. 10,
pp. 350-353, 1917.

Dobrovliansky, V. V., A list of aphids found on cultivated plants in the gov-
ernment of Kharkov, in Pests of Agriculture, Kharkov, Bull. 1916; reviewed in
Kev. Appl. Ent., vol. 5, pp. 561-562, 1917.

Essig, E. O., Aphididae of California, Univ. Calif. Publ. Entom., vol. 1, pp.
301-346, 1917.

Gillette, C. P., Some Colorado species of the genus Lachnus, Ent. Soc. Am.,
vol. 10, pp. 133-146, 1917.

Van der Goot, P., Zur Kenntnis der Blattlause Java's, in Contrib. a la fauna
der Indes neerlandaises, vol. 1, pp. 1-301, 1916.

A SYNOPSIS OF THE APHIDIDAE 179

104. Rhopalosiphum persicae (Sulzer), on Baccharis douglasii, Centranthus
ruber, Clarkia elegans, Dipsacus fullonum, Escallonia pulverulenta, Helianthus
annuus, Lavatera assurgentiflora, Liriodendron tulipifera, Melianthus major,
Pentstemon spectabilis, PittosporUm spp., and Umbellularia californica, Berkeley
(Essig, pp. 331-332).

111. Aphis avenae Fabr. It would appear from a study of Baker's paper in
Science that the common California species is Aphis prunifoliae Fitch. It is
certain that it is distinct from A. cerasifoliae Fitch, which has been taken here
once and is described in this paper. If it is possible, as Baker says, that A.
cerasifoliae Fitch is a synonym of A. padi Linn., then our common species must
be known as A. prunifoliae Fitch. From the brief description of Aphis (Siphon-
aphis) padi Linn, given by Van der Goot (pp. 71-72) it would appear that our
species may be distinct, differing slightly in the comparative lengths of the
cornicles and cauda. Consequently the author favors accepting the name, Aphis
prunifoliae Fitch, for this species.

123. Aphis gossypii Glover, on Asclepias speciosa, A. vestita, Lilium speciosum
rubrum, Lonicera sp., and Ehamnus purshiana, Berkeley and Oakland (Essig,
pp. 338-339).

131. Aphis raedicaginis Koch, on Citrus sp., Sacramento, and on Vigna sin-
ensis, Moorpark (Essig, p. 340).

139. Aphis pomi De Geer, on Cotoneaster franchetii, Pittosporum eugenioides,
and Viburnum tinus, Berkeley (Essig, p. 341). The author is inclined to believe
this to be Aphis viburnicolens n.sp. (see no. 150) which is quite similar to Aphis
pomi De Geer, but which is common on Viburnum and related plants. He has not,
however, seen Essig 's specimens, so can not state positively whether or not it is
this species.

140. Aphis prunorum Dobr. Dobrovliansky places this species as a synonym
of Siphocoryne nymphaeae (Linn.). This author noted the similarity of these
two, but was not certain of their identity, so listed them as distinct species.

141. Aphis pseudobrassicae Davis. Dobrovliansky believes this to be a syno-
nym of Aphis erysimi Kalt.

146. Aphis senecio Swain. Essig (p. 337) lists Aphis bakeri Cowen from
Trifolium pratense. This proves to be the true Aphis bakeri Cowen and not
A. senecio Swain, which is the species that has been hitherto called A. balceri
Cowen in California.

152. Toxoptera aurantii (Fonsc.) on Camellia japonica, Oakland (Essig, p.
330).

153. Hyalopterus arundinis (Fabr.). Both Dobrovliansky and Van der Goot
list this as a synonym of H. pruni (Fabr.) giving the later preference. Accord-
ing to Hunter, arundinis should have priority, but it is entirely possible that the
dates he gives are incorrect. This point the author is unable to settle as he has
not access to Fabricius' works.

156. Siphocoryne nymphaeae (Linn.). Davidson gives a brief account of the
habits and biology of this species, as well as a description of the various forms.

175. Eriosoma languinosa Hartig (pyricola Baker and Davidson). The
species listed by Essig (p. 345) as Eriosoma sp. on Ulmus campestris in Berkeley
and in Hayward is this species.

115. Aphis carl Essig. Davidson recently remarked to the author that he
could see no difference between this species and Aphis helianthii Monell. It is
quite possible that these are synonyms.

EXPLANATION OF PLATES
PLATE 1

Fig. 1. Myzocallis asclepiadis (Fitch), tarsus and claw.

Fig. 2. Aphis senecio Swain, tarsus and claw.

Fig. 3. Essigella calif ornica (Essig), sixth antennal segment and spur.

Fig. 4. Aphis senecio Swain, sixth antennal segment and spur.

Fig. 5. Essigella californica (Essig), cauda and anal plate (lateral view).

Fig. 6. Aphis senecio Swain, cauda and anal plate (lateral view).

Fig. 7. Eucallipterus tiliae (Linn.), cauda and anal plate.

Fig. 8. Thomasia populicola (Thos.), cauda and anal plate.

Fig. 9. Phyllaphis fagi (Linn.), third antennal segment.

Fig. 10. Phyllaphis fagi (Linn.), sixth antennal segment.

Fig. 11. Phyllaphis fagi (Linn.), cauad and anal plate.

Fig. 12. Phyllaphis fagi (Linn.), front of head and antennal tubercles.

Fig. 13. Phyllaphis coweni (Ckll.), Antenna.

Fig. 14. Phyllaphis quercicola Baker, third antennal segment.

Fig. 15. Phyllaphis quercicola Baker, fourth antennal segment.

Fig. 16. Phyllaphis quercicola Baker, fifth antennal segment.

Fig. 17. Phyllaphis quercicola Baker, sixth antennal segment.

Fig. 18. Phyllaphis quercicola Baker, forewing.

Fig. 19. Phyllaphis quercicola Baker, cauda and anal plate.

Fig. 20. Phyllaphis querci, tarsal claw.

Fig. 21. Drepanosiphum platanoides (Schr.), antennal tubercles.

[180]

[SWAIN ] PLATE 1

PLATE 2

Fig. 22. Myzocallis arundicolens (Clarke), antennal tubercles.

Fig. 23. Drepanaplils acerifolii (Thomas), cornicle.

Fig. 24. Drepano»iphum platanoides (Schr.), cornicle.

Fig. 25. Monellia caryella (Fitch), cornicle.

Fig. 26. Myzocallis bcllus (Walsh), cornicle.

Fig. 27. Calaphis betulaecolens (Fitch), antennal tubercles.

Fig. 28. Calaphis betulella Walsh, antennal tubercles.

Fig. 29. Euceraphis betulae (Koch), antennal tubercles.

Fig. 30. Eucallipterus tiliae (Linn.), sixth antennal segment and spur.

Fig. 31. Myzocallis quercus (Kalt.), sixth antennal segment and spur.

Fig. 32. Myzocallis quercus (Kalt.), cornicle.

Fig. 33. Eucallipterus tiliae (Linn.), cornicle.

Fig. 34. Chromaphis juglandicola (Kalt.), sixth antennal segment and spur.

Fig. 35. Chromaphis juglandicola (Kalt.), cornicle.

Fig. 36. Drepanosiphum platanoides (Schr.), third antennal segment.

Fig. 37. Drepanaphis acerifolii (Thomas), third antennal segment.

Fig. 38. Calaphis betulae-colens (Fitch), third antennal segment.

Fig. 39. Euceraphis gillettei Dvdn., base of third antennal segment.

Fig. 40. Euceraphis betulae (Koch), base of third antennal segment.

[ SWAIN ] PLATE 2

PLATE 3

Fig. 41. Eucalli'pterus flava (Dvdn.), base of third antennal segment.

Fig. 42. Eucallipterus tiliae (Linn.), third antennal segment.

Fig. 43. Myzocallis coryli (Goetze), third antennal segment.

Fig. 44. Myzocallis coryli (Goetze), sixth antennal segment and spur.

Fig. 45. Myzocallis bellm (Walsh), sixth antennal segment and spur.

Fig. 46. Myzocallis bellus (Walsh), third antennal segment.

Fig. 47. Myzocallis alnifoliae (Fitch), third antennal segment.

Fig. 48. Myzocallis arundicolens (Clarke), third antennal segment. ,

Fig. 49. Eucallipterus tiliae (Linn.), cornicle.

Fig. 50. Eucallipterus tiliae (Linn.), anal plate.

Fig. 51. Myzocallis arundicolens (Clarke), cornicle.

Fig. 52. Myzocallis arundicolens (Clarke), anal plate.

Fig. 53. Myzocallis coryli (Goetze), cornicle.

Fig. 54. Myzocallis coryli (Goetze), anal plate.

Fig. 55. Myzocallis californicus Baker, third antennal segment.

Fig. 56. Myzocallis californicus Baker, sixth antennal segment and spur.

Fig. 57. Myzocallis pasaniae Dvdn., third antennal segment.

Fig. 58. Myzocallis quercus (Kalt.), third antennal segment.

Fig. 59. Myzocallis ulmifolii (Monell), third antennal segment.

Fig. 60. Myzocallis castanicola Baker, third antennal segment.

Fig. 61. Myzocallis castanicola Baker, cauda and anal plate.

Fig. 62. Myzocallis castanicola Baker, cornicle.

Fig. 63. Callipterus californicus (Essig), sixth antennal segment and spur.

Fig. 64. Callipterus californicus (Essig), third antennal segment.

Fig. 65. Callipterus caryae Monell, third antennal segment.

[184]

5a3 TS

49

50

55
56

57

5\

54

[ SWAIN ] PLATE 3

PLATE 4

Fig. 66. Callipterus caryae Monell, sixth antennal segment and spur.

Fig. 67. Monellia caryella (Fitch), sixth antennal segment and spur.

Fig. 68. Monellia caryella (Fitch), third antennal segment.

Fig. 69. Arctaphis populifolii (Essig), cauda.

Fig. 70. Micrella monella Essig, cauda.

Fig. 71. Arctaphis populifolii (Essig), third antennal segment.

Fig. 72. Micrella monella Essig, third antennal segment.

Fig. 73. Symydobius macrostachyae Essig, third antennal segment.

Fig. 74. Symydobius salicicorticis Essig, third antennal segment.

Fig. 75. Fullawaya saliciradicis Essig, third antennal segment.

Fig. 76. Thomasia, crucis Essig, third antennal segment.

Fig. 77. Thomasia populicola (Thomas), third antennal segment.

Fig. 78. Thomasia salicicola (Essig), third antennal segment.

Fig. 79. Lachnus ferrisi Swain, tarsal claw.

Fig. 80. Pterocomma populifoliae (Fitch), tarsal claw.

Fig. 81. Pterocomma flocculosa (Weed), cornicle.

Fig. 82. Pterocomma populifoliae (Fitch), cornicle.

Fig. 83. Essigella calif ornica (Essig), antenna.

Fig. 84. Longistigma sp., front wing.

[186]

OQ06

71

70

74

[ SWAIN ] PLATE 4

PLATE 5

Fig. 85. Lachnus sp., front wing.

Fig. 86. Tuberolachnus viminalis (Fonsc.), hind tarsus.

Fig. 87. Eulachnus rileyi Davis, hind tarsus.

Fig. 88. Lachnus vanduzei n.sp., third antennal segment.

Fig. 89. Lachnus ferrisi Swain, first, second, and third antennal segments.

Fig. 90. Laohnus ferrisi Swain, fourth, fifth, and sixth antennal segments.

Fig. 91. Lachnus ferrisi Swain, cornicle.

Fig. 92. Lachnus pseudotsugae Wilson, tip of front wing.

Fig. 93. Lachnus tujafilinus (Del Guercio), tip of front wing.

Fig. 94. Lachnus occidentalis Dvdn., third antennal segment.

Fig. 95. Lachnus pini-radiatae Dvdn. (?), third antennal segment.

Fig. 96. Lachnus glehnus Essig, third antennal segment.

Fig. 97. Lachnus glehnus Essig, cornicle.

Fig. 98. Lachnus pseudotsugae Wilson, third antennal segment.

Fig. 99. Lachnus taxifolia Swain, hind tarsus.

Fig. 100. Lachnus taxifolia Swain, fourth, fifth and sixth antennal segments.

Fig. 101. Lachnus taxifolia Swain, first, second, and third antennal segments.

[188]

[ SWAIN 1 PLATE 5

PLATE 6

Fig. 102. Laolmus taxifolia Swain, wing.

Fig. 103. Lachnus taxifolia Swain, cornicle of apterous female.

Fig. 104. Lachnus ponderosa Williams, third antennal segment.

Fig. 105. Lachnus tujafilinus (Del Guercio), third antennal segment.

Fig. 106. Macrosiphum rosae (Linn.), antennal tubercles.

Fig. 107. Nectaro-siphon rubicola (Oest.), antennal tubercles.

Fig. 108. Ehopalosiphum persicae (Sulz.), antennal tubercles.

Fig. 109. Nectarosiphon rubicola (Oest.), cornicle.

Fig. 110. Idiopterus nephrelepidis Davis, wing.

Fig. 111. Amphorophora rubi (Kalt.), antennal tubercles.

Fig. 112. Myzus cerasi (Fabr.), antennal tubercles.

Fig. 113. Amplioropliora rubi (Kalt.), cornicle.

Fig. 114. Toxoptera aurantii (Fonsc.), cornicle.

Fig. 115. Phorodon Jiumuli (Schr.), antennal tubercles of alate females.

Fig. 116. Phorodon Jiumuli (Schr.), antennal tubercles of apterous females.

UH>|

113

[SWAIN] PLATE 6

PLATE 7

Fig. 117. Phorodon humuli Schr., cornicle.

Fig. 118. Phorodon humuli Schr., cauda.

Fig. 119. Ehopalosiphum persicae (Sulz.), cornicle.

Fig. 120. Ehopalosiphum pcsicae (Sulz.), cauda.

Fig. 121. Myzus cerasi (Fabr.), cornicle.

Fig. 122. Myzus cerasi (Fabr.), cauda.

Fig. 123. Nectarosiphon rubicola (Oest.), cauda.

Fig. 124. Nectarosiphon morrisoni Swain, antennal tubercles.

Fig. 125. Nectarosiphon morrisoni Swain, third antennal segment.

Fig. 126. Nectarosiphon morrisoni Swain, cauda.

Fig. 127. Nectarosiphon morrisoni Swain, cornicle.

Fig. 128. Macrosiphum stanleyi Wilson, cornicle.

Fig. 129. Macrosiphum solanifolii (Ashm.) (from Sonchus), cornicle.

Fig. 130. Macrosiphum pisi (Kalt.), cornicle.

Fig. 131. Macrosiphum californicum (Clarke), third antennal segment.

Fig. 132. Macrosiphum californicum (Clarke), cornicle.

Fig. 133. Macrosiphum cucurbitae (Thomas), third antennal segment.

Fig. 134. Macrosiphum cucurbitae (Thomas), cornicle.

Fig. 135. Macrosiphum granarium (Kirby), third antennal segment.

Fig. 136. Macrosiphum ludovicianae (Oest.), third antennal segment.

Fig. 137. Macrosiphum solanifolii (Ashm.), cornicle.

Fig. 138. Macrosiphum solanifolii (Ashm.), third antennal segment.

[192]

SoS? 0°°o ^ °o °0 Qo°o° 'o O O go°o g 00000o0o°

I3JT

o o o ° o o OQ OQ_00__; 3 go °

[ SWAIN ] PLATE 7

PLATE 8

Fig. 139. Macrosiphum solanifolii (Ashm.) (from Citrus), cornicle.

Fig. 140. Macrosiphum solanifolii (Ashm.) (from Citrus), third antennal
segment.

Fig. 141. Macrosiphum sanborni Gillette, cornicle of apterous female.

Fig. 142. Macrosiphum artemisiae (Fonsc.), cornicle.

Fig. 143. Macrosiphum albifrons Essig, third antennal segment.

Fig. 144. Macrosiphum albifrons Essig, cornicle.

Fig. 145. Macrosiphum artemisiae (Fonsc.), third antennal segment.

Fig. 146. Macrosiphum artemisicola (Williams), third antenal segment.

Fig. 147. Macrosiphum artemisicola (Williams), cornicle.

Fig. 148. Macrosiphum granarium (Kirby), cornicle.

Fig. 149. Macrosiphum ludovicianae (Oest.), cornicle.

Fig. 150. Macrosiphum pisi (Kalt.), third antennal segment.

Fig. 151. Macrosiphum rosae (Linn.), third antennal segment.

Fig. 152. Macrosiphum rosae (Linn.), cornicle.

Fig. 153. Macrosiphum rudbeclciae (Fitch), cornicle.

Fig. 154. Macrosiphum rudbeclciae (Fitch), third antennal segment.

Fig. 155. Macrosiphum sanborni Gillette, cauda apterous female.

Fig. 156. Macrosiphum dirhodum (Walker), cornicle.

Fig. 157. Macrosiphum dirhodum (Walker), third antennal segment.

Fig. 158. Macrosiphum stanleyi Wilson, third antennal segment.

Fig. 159. Macrosiphum solanifolii (Ashm.) (from Souchus), third antennal
segment.

Fig. 160. Macrosiphum solanifolii (Ashm.) (from Sonchus), cauda.

194]

160

[SWAIN ] PLATE 8

Fig. 161.

Fig. 162.

Fig. 163.

Fig. 164.

Fig. 165.

Fig. 166.

Fig. 167.

Fig. 168.

Fig. 169.
segment.

Fig. 170.

Fig. 171.
ment.

Fig. 172.

Fig. 173.

Fig. 174.

Fig. 175.

Fig. 176.

Fig. 177.

Fig. 178.

Fig. 179.

Fig. 180.

Fig. 181.

Fig. 182.

Fig. 183.

Fig. 184.

Fig. 185.
ments.

Fig. 186.

PLATE 9

Amphorophora latysiphon Dvdn., cornicle.

Amphorophora rubi (Kalt.), cauda.

Toxoptera aurantii (Fonsc.), third antennal segment.

Ehopalosiphum violae Pergande, wing.

Ehopalosiphum hippophaes Koch, cornicle.

Ehopalosiphum nervatum Gillette (from Arbutus), wing.

Ehopalosiphum corylinum Dvdn., third antenpal segment.

Ehopalosiphum persicae (Sulz.), third antennal segment.

Ehopalosiphum nervatum Gillette (from Arbutus), third antennal

Bhopalosiphum hippophaes Koch, third antennal segment.
Ehopalosiphum nervatum Gillette (from rose), third antennal seg-

Siphocoryne nymphaeae (Linn.), third antennal segment.

Ehopalosiphum rhois Monell, third antennal segment.

Ehopalosiphum violae Pergande, third antennal segment.

Myzus circumflexus (Buckton), third antennal segment.

Myzus braggii Gillette, third antennal segment.

Myzus fragaefolii Ckll., third antennal segment.

Myzus rhamni (Fonsc.), third antennal segment.

Myzus cerasi (Fabr.), third antennal segment.

Myzus ribis (Linn.), third antennal segment.

Hyalopterus arundinis (Fabr.), cornicle.

Aphis euonomi Fabr., cornicle.

Siphocoryne capreae (Fabr.), cornicle.

Liosomaphis berberidis (Kalt.), conricle.

Hyalopterus arundinis (Fabr.), third and fourth antennal seg-

Hyalopterus arundinis (Fabr.), cauda.

[196]

IS2

if 5

( SWAIN ] PLATE 9

PLATE 10

Fig. 187. Aphis euonomi Fabr., wing.

Fig. 188. Aphis salicicola Thomas, wing.

Fig. 189. Aphis medicaginis Koch, third and fourth antennal segments.

Fig. 190. Aphis euonomi Fabr. (I), third and fourth antennal segments.

Fig. 191. Aphis avenae Fabr., wing.

Fig. 192. Aphis gossypii Glover, cornicle.

Fig. 193. Aphis gossypii Glover, cauda.

Fig. 194. Aphis sambucifoliae Fitch, cauda.

Fig. 195. Aphis sambucifoliae Fitch, cornicle.

Fig. 196. Myzaphis abietina (Walker), third and fourth antennal segments.

Fig. 197. Myzaphis abietina (Walker), cornicle.

Fig. 198. Aphis albipes Oest., cornicle.

Fig. 199. Aphis albipes Oest., cauda.

Fig. 200. Aphis albipes Oest., third and fourth antennal segments.

Fig. 201. Aphis avenae Fabr., cornicle.

Fig. 202. Aphis avenae Fabr., third and fourth antennal segments.

Fig. 203. Aphis brassicae Linn., cornicle.

Fig. 204. Aphis brassicae Linn., third and fourth antennal segments.

Fig. 205. Aphis euonomi Fabr., cornicle.

Fig. 206. Aphis euonomi Fabr., cornicle.

Fig. 207. Aphis euonomi Fabr., third and fourth antennal segments.

Fig. 208. Aphis cardui Linn., third and fourth antennal segments.

Fig. 209. Aphis cardui Linn., cornicle.

Fig. 210. Aphis ceanothi Clarke, cornicle.

Fig. 211. Aphis ceanothi Clarke, third and fourth antennal segments.

Fig. 212. Aphis coolcii Essig, third and fourth antennal segments.

Fig. 213. Aphis coolcii Essig, cauda and anal plate.

Fig. 214. Aphis coolcii Essig, cornicle.

Fig. 215. Aphis gossypii Glover, third and fourth antennal segments.

Fig. 216. Aphis maidis Fitch, cauda.

Fig. 217. Aphis maidis Fitch, antenna.

Fig. 218. Aphis maidis Fitch, cornicle.

Fig. 219. Aphis middletonii Thomas, cornicle.

Fig. 220. Aphis middletonii Thomas, third and fourth antennal segments.

Fig. 221. Aphis nerii Fonsc., cornicle.

Fig. 222. Aphis nerii Fonsc., third and fourth antennal segments.

Fig. 223. Aphis persicae-niger Smith, cornicle.

Fig. 224. Aphis persicae-niger Smith, third and fourth antennal segments.

[198]

I SWAIN ] PLATE 10

PLATE 11

Aphis pomi De Geer, canda.

Aphis pomi De Geer, antennae.

Aphis pomi De Geer, cornicle.

Aphis prunorum Dobr., cauda.

Aphis prunorum Dobr., third and fourth antennal segments.

Aphis prunorum Dobr., cornicle.

Aphis pscudobrassicac Davis, third and fourth antennal segments.

Aphis ramona Swain, antenna.

Aphis ramona Swain, front of head.

Aphis ramona Swain, cauda and anal plate.

Aphis ramona Swain, cornicle.

Aphis euonomi Fabr., cornicle.

Aphis euonomi Fabr., third and fourth antennal segments.

Aphis salicicola Thomas, cornicle.

Aphis salicicola Thomas, third and fourth antennal segments.

Aphis sambucifoliae Fitch, third and fourth antennal segments.

Aphis senecio Swain, cauda.

Aphis senecio Swain, cornicle.

Aphis senecio Swain, front of head.

Aphi-s senecio Swain, third and fourth antennal segments.

Aphis senecio Swain, fifth, sixth antennal segments, and spur.

Aphis setarae Thomas, cornicle.

Aphis setarae Thomas, third and fourth antennal segments.

Aphis malifoliae Fitch, cornicle.

Aphis malifoliae Fitch, fourth antennal segment.

Liosomaphis berberidis (Kalt.), front of head.

Liosomaphis berberidis (Kalt.), third and fourth antennal seg-

Siphocoryne capreae (Fabr.), third and fourth antennal segments.
Siphocoryne capreae (Fabr.), fifth and sixth antennal segments

Siphocoryne capreae (Fabr.), cauda and supra-caudal spine of alate

Fig. 225.

Fig. 226.

Fig. 227.

Fig. 228.

Fig. 229.

Fig. 230.

Fig. 231.

Fig. ( 232.

Fig. '233.

Fig. 234.

Fig. 235.

Fig. 236.

Fig. 237.

Fig. 238.

Fig. 239.

Fig. 240.

Fig. 241.

Fig. 242.

Fig. 243.

Fig. 244.

Fig. 245.

Fig. 246.

Fig. 247.

Fig. 248.

Fig. 250.

Fig. 251.

Fig. 252.
ments.

Fig. 253.

Fig. 254.
and spur.

Fig. 255.
females.

Fig. 256. Siphocoryne capreae (Fabr.), cauda and supra-caudal spine of
apterous females.

Fig. 257. Siphocoryne pastinacae (Linn.), third and fourth antennal seg-
ments.

Fig. 258. Siphocoryne pastinacae (Linn.), fifth and sixth antennal segments
and spur.

Fig. 259. Siphocoryne pastinacae (Linn.), cauda of apterous female.

Fig. 260. Siphocoryne pastinacae (Linn.), cauda of alate female.

Fig. 261. Siphocoryne pastinacae (Linn.), cornicle.

[200]

241

[ SWAIN ] PLATE 1 1

PLATE 12

Fig. 262. Mysocallis discolor (Monell), fore wing.

Fig. 263. Myzocallis discolor (Monell), third antennal segment.

Fig. 264. Mysocallis bellus (Walsh), fore wing.

Fig. 265. Myzocallis bellus (Walsh), third antennal segment.

Fig. 266. Myzocallis calif ornicus Baker (maureri Swain), fore wing.

Fig. 267. Myzooallis castanicola Baker ^davidsoni Swain), fore wing.

Fig. 268. Myzocallis arundinariae Essig, third antennal segment.

[202]

2 66

ztr

[ SWAIN ] PLATE 12

~(j U v UV U — •— •f'

PLATE 13

Fig. 269. Symydobius chrysolepis Swain, head.

Fig. 270. Symydobius chrysolepis Swain, cornicle.

Fig. 271. Symydobius chrysolepis Swain, anal plate.

Fig. 272. Symydobius chrysolepis Swain, antenna.

Fig. 273. Symydobius chrysolepis Swain, fore wing.

Fig. 274. Symydobius chrysolepis Swain, hind wing.

Fig. 275. Thomasia populicola (Thomas), fore wing.

[204]

Z72

274

275

[ SWAIN ] PLATE 13

PLATE 14

Fig. 276. Toxoptera aurantii (Fonsc.), fore wing.

Fig. 277. Rhopalosiphum lactucae (Kalt) head.

Fig. 278. Rhopalosiphum lactucae (Kalt.), third antennal segment, aptera.

Fig. 279. Rhopalosiphum lactucae (Kalt.), third antennal segment, alate.

Fig. 280. Rhopalosiphum lactucae (Kalt.), fourth and fifth antennal seg-
ments, alate.

Fig. 281. Rhopalosiphum lactucae (Kalt.), sixth antennal segment, alate.

Fig. 282. Rhopalosiphum lactucae (Kalt.), cornicle, alate.

Fig. 283. Rhopalosiphum lactucae (Kalt.), cauda, alate.

Fig. 284. Rhopalosiphum lactucae (Kalt.), cornicle, aptera.

Fig. 284a. Rhopalosiphum lactucae (Kalt.), cauda, aptera.

[206]

2 7k

277

Z&Z

2Z4

279

2 SO

[SWAIN] PLATE 14

PLATE 15

Fig. 285. Aphis viburnicolens n.sp., third antennal segment.

Fig. 286. Aphis viburnicolens n.sp., cornicle.

Fig. 287. Aphis viburnicolens n.sp., cauda.

Fig. 288. Aphis cerasifoliae (Fitch), head.

Fig. 289. Aphis cerasifoliae (Fitch), fifth and sixth antennal segments.

Fig. 290. Aphis cerasifoliae (Fitch), third and fourth antennal segments.

Fig. 291. Aphis cerasifoliae (Fitch), end of wing.

Fig. 292. Aphis cerasifoliae (Fitch), side of abdomen showing cauda, cor-
nicle, and lateral tubercles on segments one, two, three, four, and seven.

[208]

2S5

[ SWAIN ] PLATE 15

Tig. 293. Aphis

Fig. 294. Aphis

Fig. 295. Aphis

Fig. 296. Aphis

Fig. 297. Aphis

Fig. 298. Aphis

Fig. 299. Aphis

Fig. 300. Aphis
antennal segments.

Fig. 301. Aphis

Fig. 302. Aphis

Fig. 303. Aphis

Fig. 304. Aphis

Fig. 305. Aphis

PLATE 16

marutae Oest., head.

marutae Oest., third and fourth antennal segments.

marutae Oest., fifth and sixth antennal segments.

marutae Oest., antenna, aptera.

marutae Oest., end of abdomen, aptera.

marutae Oest., cauda, alate.

marutae Oest., cornicle, alate.

neomexicana Ckll., var. paoifica Dvdn., third and fourth

neomexicana Ckll. var. pacifica Dvdn., cornicle.

neomexicana Ckll. var. pacifica Dvdn., cauda.

yuccae Cowen, fourth, fifth, and sixth antennal segments.

yuccae Cowen, third antennal segment.

yuccae Cowen, tip of abdomen.

293

294

295

296

29S-

299

do;

30*

504

SWAIN ] PLATE; 16

PLATE 17

Fig. 306. Myzus ribis (Linn.), head.

Fig. 307. Myzus cerasi (Fabr.), head.

Fig. 308. Myzaphis rosarum (Walker), head, alate.

Fig. 309. Myzaphis rosarum (Walker), third and fourth antennal segments.

Fig. 310. Myzaphis rosarum (Walker), fifth and sixth antennal segments.

Fig. 311. Mysaphis rosarum (Walker), tip of wing.

Fig. 312. Myzaphus rosarum (Walker), end of abdomen.

Fig. 313. Myzaphis rosarum (Walker), head, aptera.

Fig. 314. Myzaphis rosarum (Walker), antenna, aptera.

Fig. 315. Myzaphis rosarum (Walker), cornicle, aptera.

Fig. 316. Myzaphis rosarum (Walker), cauda, aptera.

Fig. 317. Myzaphis rosarum (Walker), hind tarsus, aptera.

[212]

507

303"

31 3

315

[ SWAIN 1 PLATE 17

A SYNOPSIS OF THE APHIDIDAE

215

INDEX TO GENEEA AND SPECIES

abietes, Lachnus, 47.
abietina, Myzaphis (Aphis), 134.
abietinus, Mindarus, 150.
acerifolii, Drepanaphis (Siphono-

phora, Macrosiphum), 18.
achyrantes, Rhopalosiphum, 80.
agrifoliae, Symydobius, 38.
alamedensis, Aphis, 93.
albifrons, Macrosiphum, 60.
albipes, Aphis, 93.
alni, Myzocallis, 21.
alnifoliae, Callipterus, 20.
alnifoliae Lachnus, 20.
alnifoliae Myzocallis, 22.
alnifoliae Prociphilus (Pemphigus},

146.
ambrosiae, Macrosiphum (Siphono-

phora), 60.
americana, Eriosoma (Schizoneura),

148.
Amphorophora, 54.

cicutae, 54.

latysiphon, 54, 178.

rubi, 54.

rubicola, 77.
angelicae, Aphis, 93.
annulata, -Callipterinella (Chaitopho-

rus), 31.
Aphis, 88.

abietina, 134.

alamedensis, 93.

albipes, 93.

angelicae, 93.

artemisiae, 61.

arundinis, 130.

atriplicis, 93.

aurantii, 129.

avenae, 94, 179.

bakeri, 123, 124.

bakeri, 6, 179.

h< 'I in*. 24.

berberidis, 130.

betulaecolens, 18.

brassicae, 95.

calendulicola, 96.

capreae, 132.

cardui, 96.

cari, 96, 179.

caryella, 30.

ceanothi, 96.

ceanothi-hirsuti, 96.

cerasi, 73.

cerasifoliae, 97.

citri, 105.

eooki, 100.

cornifoliae, 100.

coryli, 25.

crataegifoliae, 100.

dirhodum, 63.
dryophila, 150.
euonomi, 101.
fabae, 102, 104.
fagi, 13.
frigidae, 105.
gossypii, 100.
gossypii, 105, 179.
granarium, 64.
hederae, 106.
heraclei, 107.
houghtonensis, 107.
humuli, 79.
juglandis, 28.
lactucae, 82.
languinosa, 149.
lanigerum, 149.
lithospermi, 108.
lutescens, 117.
maidis, 94.
maidis, 108.
mali, 120.
malifoliae, 108.
marutae, 112.
medicaginis, 114, 179.
middletonii, 115.
mori, 116.
neomexicana, 116.
nerii, 117.
nymphaeae, 133.
oenotherae, 118.
oregonensis, 119.
padi, 94.

papaveris, 102, 104.
pastinacae, 133.
persicae, 85.
persicae-niger, 119.
pisi, 66.
platanoides, 17.
pomi, 120, 179.
pomi, 109.
populifoliae, 41.
pruni, 96.

prunorum, 121, 179.
prunifoliae, 130, 179.
pseudobrassicae, 122, 179.
quercus, 27.
ramona, 122.
rhamni, 76.
rosae, 67.
rosarum, 134.
rubi, 54.
rubiphila, 122.
rudbeckiae, 67.
rufomaculata, 137.
rumtcis, 101, 106.
salicicola, 123.
sambucifoliae, 123.
senecio, 123, 179.
setariae, 124.

216

MISCELLANEOUS STUDIES

sorbi, 108.

spiraecola, 1-1.

spiraeella, 125, 126.

taraxici, 71.

tetrapteralis, 125.

tiliae, 21.

viburnicolens, 126, 179.

viminalis, 45.

yuccae, 45.

yuccicola, 128.
aquilegiae, Myzus, 73.
arbuti, Ehopalosiphum, 84.
Aretaphis, 33.

populifolii, 33.

viminalis, 34.
artemisicola, Macrosiphum (Siphono-

phora), 61.

artemisiae, Macrosiphum (Aphis), 61.
arundicolens, Eucallipterus (Myzocal-
lis}, 24.
arundicolens, Myzocallis (Callipterus) ,

22.

arundinariae, Myzocallis, 24.
arundinis, Hyalopterus (Aphis'), 130,

179.

atriplicis, Aphis, 93.
aurantiae, Toxoptera, 129.
aurantii, Toxoptera (Aphis), 129, 179.
avenae, Aphis (Nectarophora, Sipho-
coryne), 94, 179.

B

baccharadis, Macrosiphum (Nectaro-
phora), 61.

bakeri, Aphis, 123.

bakeri, Aphis, 6, 179.

balsamiferae, Pemphigus, 142.

bellus, Myzocallis (Aphis, Callip-
terus), 24.

berberidis, Liosomaphis (Aphis, Bho-
palosiphum) , 130.

betae, Pemphigus, 142.

betulae, Chaitophorus, 31.

betulae, Euceraphis (Callipterus) , 19,
178.

betulaecolens, Calaphis (Aphis, Cal-
lipterus), 18.

braggii, Myzus, 73.

brassicae, Aphis, 95.

Byrsocrypta, 148.
ulmicola, 148.

calendulieola, Aphis, 96.

Calaphis, 18.

betulaecolens, 18.
castaneae, 24.

calif ornica, Essigella (Lachnus), 44.

calif ornicum, Macrosiphum (Nectaro-
phora), 62.

calif ornicus, Callipterus (Monellia),
29.

californicus Myzocallis, 178.

calif ornicus Thecabius (Pemphigus),
144.

Callipterinella, 31.

annulata, 31.
Callipterus, 28.

alnifoliae, 20.

arundicolens, 22.

bellus, 24.

betulae, 19.

betulaecolens, 18.

californicus, 29.

caryae, 29.

caryella, 30.

castaneae, 24.

coryli, 25.

discolor, 25.

hyalinus, 26.

juglandicola, 28.

juglandis, 28.

punctatus, 26.

quercus, 27.

tiliae, 21.

ulmifolti, 27.

viminalis, 34.

capreae, Siphocoryne (Aphis), 132.
cardui, Aphis, 96.
carduinum, Phorodon, 73.
cari, Aphis, 96, 179.
caryae, Callipterus (ifone^ia), 29.
caryella, Monellia ( Aphis, Callip-
terus), 30.

castaneae, Calaphis (Callipterus), 24.
castaneae, Myzocallis, 178.
castanicola, Myzocallis, 178.
ceanothi, Aphis, 96.
ceanothi-hirsuti, Aphis, 96.
cerasi, Myzus (Aphis), 73.
cerasifoliae, Aphis, 97.
Cerataphis, 140.

lataniae, 140.
Cerosipha, 137.

cupressi, 137.
Chaitophorus, 33.

annulata, 31.

betulae, 31.

negundinis, 36.

nigrae, 37.

populieola, 36.

populifoliae, 33.

salicieola, 37.

smithiae, 34.

vtmt'jiaZis, 34.
Chermes, 151.

cooleyi, 151.

coweni, 151.

pinicorticis, 152.
Chromaphis, 28.

juglandicola, 28.
chrysanthemi, Macrosiphum (Siphono-

phora), 62.

chrysanthemi, Macrosiphum, 69.
chrysolepis, Symydobius, 38.
cicutae, Amphorophora, 54.
circumflexus, Myzus (Siphonophora).
citri, Aphis, 105.

A SYNOPSIS OF THE APHIDIDAE

217

citrifolii, Macrosiphum (Nectaro-

phora), 69.
Cladobius, 41.

rufulus, 41.

salicti, 43.
Coccus, 140.

lataniac, 140.

pinicorticis, 152.
Colopha, 148.

ulmicola, 148.
Coloradoa, 137.

rufomaculata, 137.
conii, Siphocoryne, 133.
cooki, Aphis, 100.
cooleyi, Chermes, 151.
cornifoliae, Aphis, 100.
coryli, Myzocallis (Aphis, Callip-
terus), 25.

corylinum, Rhopalosiphum, 81.
coweni, Chermes, 151.
coweni, Phyllaphis (Pemphigus), 13.
crataegifolii, Aphis, 100.
crueis, Thomasia, 36.
Cryptosiphum, 13.

tahoense, 13.
cucurbitae, Macrosiphum (Siphono-

phora), 62.

cupressi, Cerosipha, 137.
cynosbati, Myzus (Nectarophora), 75.

D

davidsoni, Myzocallis, 24, 178.
dentatus, Laclmus, 45.
destructor, Macrosiphum, 66.
dianthi, Rhopalosiphum, 85.
dirhodum, Macrosiphum (Aphis), 63.
discolor, Myzocallis (Callipterus) , 25.
Drepanaphis, 18.

acerifolii, 18.
Drepanosiphum, 17.

acerifolii, 18.

platanoides, 17.

dryophila, Vacuna (Aphis, Chaito-
phorus), 150.

E
Eichocliaitophorus, 33.

populifolii, 33.
Eriosoma, 148.

americana, 148.

languinosa, 149, 179.

lanigerum, 149.

pyricola, 149.
Essigella, 44.

californica, 44.
essigi, Myzocollis, 27.
Eucallipterus, 20.

arundicolens, 24.

flava, 20.

tiliae, 21, 178.
Euceraphis, 19.

betulae, 19, 178.

flava, 20.

gillettei, 20.
euonomi, Aphis, 101.

fagi, Phyllaphis, 13, 178.

ferrisi, Laclmus, 47.

flava, Eucallipterus (Euceraphis), 20.

flocculosa, Pterocomma (Melanoxan-
thus), 40.

foeniculi, SipJwcoryne, 132.

fragaefolii, Myzus, 75.

fraxini-dipetalae, Prociphilus (Pem-
phigus), 146.

frigidae, Aphis, 105.

frigidae, Macrosiphum, 61.

Fullawaya, 35.
saliciradicis, 35.

G

galeopsidis, Phorodon, 81.
gillettei, Eueeraphis, 20.
glehnus, Lachnus, 47.
godetiae, Myzus, 85.
gossypii, Aphis, 100.
gossypii, Aphis, 105, 179.
granarium, Macrosiphum (Aphis), 64,
178.

H

hederae, Aphis, 106.
heraclei, Aphis, 107.
heucherae, Macrosiphum (Sipliono-

phora), 64.

hippophoaes, Rhopalosiphum, 81.
houghtonensis, Aphis, 107.
howardi, Ehopalosiphum, 86.
humuli, Phorodon (Aphis), 79.
Hyadaphis, 132.

pastinacae, 132.

umbellulariae, 133.

hyalinus, Myzocallis (Callipterus), 26.
Hyalopterus, 130.

arundinis, 130, 179.

Idiopterus, 56.
nephrelepidis, 56.

jasmini, Macrosiphum (Nectaro-

phora), 64.
juglandicola, Chromaphis (Lachnus,

Callipterus), 28.

juglandis, Callipterus (Aphis), 28.
junipcri, Lachnus, 50.

L
Lachniella, 50.

tujafilinus, 50.
Lachnus, 45.

abietis, 47.

alnifoliae, 20, 22.

calif ornicus, 44.

dentatfts, 45.

ferrisi, 47.

glehnus, 47.

juglandicola, 28.

juniperi, 50.

218

MISCELLANEOUS STUDIES

occidentalis, 17.

oregonensis, 48.

pini-radiatae, 48, 178.

ponderosa, 48.

pseudotsugae, 48.

sabinianus, 49.

taxifolia, 50.

tomentosus, 178.

tujafilinus, 50.

vanduzei, 50.

viminalis, 45.

lactucae, Macrosiphum (Nectaro-
phora), 65.

lactuca Rhopalosiphum (Aphis), 82.
laevigatae, Macrosiphum, 62.
languinosa, Eriosoma (Aphis), 149,

179.
langerum, Eriosoma (Aphis, Schiso-

neura), 149.

lataniae, Cerataphis (Coccus), 140.
latysiphon, Amphorophora, 54, 178.
Liosomaphis, 130.

berberidis, 130.
lithospermi, Aphis, 108.
ludovieianae, Macrosiphum (Siplwno-

phora), 65.
lutescens, Aphis, 117.
lycopersici, Myzus (Nectarophora),
76.

M

Macrosiphum, 57.
acerifolii, 18.
albifrons, 60.
ambrosiae, 60.
artemisiae, 61.
artemisicola, 61.
baccharadis, 61.
californicum, 62.
chrysanthemi, 62.
chrysanthemi, 69.
citrifolii, 69.
cucurbitae, 62.
destructor, 66.
dirhodum, 63.
frigidae, 61.
granarium, 64, 178.
heucherae, 64.
jasmini, 64.
lactucae, 65.
laevigatae, 62.
ludovieianae, 65.
orthocarpus, 66.
pisi, 66.
pteridis, 67.
rosae, 67.
rubicola, 77.
rudbeckiae, 67.
rudbeckiae var. madia, 68.
sanborni, 69.
solanifolii, 69, 178.
sonchella, 70.
sonchi, 60.
stanleyi, 70.

taraxici, 71.

tulipae, 71.

valerianae, 71.

macrostachyae, Symydobius, 38.
madia, Macrosiphum (rudbeckiae), 68.
maidis, Aphis, 84.
maidis, Aphis, 108.
mali, Aphis, 120.
malifoliae, Aphis, 108.
marutae, Aphis, 112.
maureri, Myzocallis, 26, 178.
medicaginis, Aphis, 114, 179.
Melanoxantherium, 41.

rufulum, 41.

nu.li.i-ti. 43.
Melanoxantltus, 40.

flocculosa, 40.
Micrella, 35.

monella, 35.

middletonii, Aphis, 115.
Mindarus, 150.

abietinus, 150.
monella, Micrella, 35.
Monellia, 29.

californicus, 29.

caryae, 29.

earyella, 30.
mori, Aphis, 116.
morrisoni, Ne,ctarosiphon, 78.
Myzaphis, 134.

abietina, 134.

rosarum, 134.
Myzocallis, 21.

alni, 21.

alnifoliae, 22.

arundicolens, 22.

arundicolens, 24.

arundinariae, 24.

bellus, 24.

californicus, 178.

castaneae, 178.

castanicola, 178.

coryli, 25.

davidsoni, 24, 178.

discolor, 25.

essigi, 27.

hyalinus, 26.

maureri, 26, 178.

pasaniae, 26.

punctatus, 26.

quercus, 27.

ulmifolii, 27.

woodu'orthi, 27.
Myzus, 71.

aquelegiae, 73.

braggii, 73.

cerasi, 73.

circumflexus, 74.

cynosbati, 75.

fragaefolii, 75.

godetiae, 85.

lycopersici, 76.

persicae, 80, 85.

rhamni, 76.

A SYNOPSIS OF THE APHIDIDAE

219

ribes, 75.
ribifolii, 76.
rosarum, 134.
varians, 77.
mncae, 74.

N

Nectarophora

averiae, 94.

bacoharadis, 61.

californicum, 62.

c-itrifolii, 69.

oynosbati, 75.

jasmini, 64.

lactucae, 65.

lycopersici, 76.

pm, 66.

rhamni, 76.

?-osae, 67.

soncliella, 70.

valerianae, 71.
Nectarosiphon, 77.

morrisoni, 78.

rubicola, 77.
negundinis, Thomasia (Chaitophorus) ,

36.

neomexicana, Aphis, 116.
nephrelepidis, Idiopterus, 56.
nerii, Aphis, 117.
nervatum, Khopalosiphum, 84.
nigrae, Chaitophorus, 37.
nigronervoaa, Pentalonia, 78.
nymphaeae, Siphocoryne (Aphis, Eho-
palosiphum), 133, 179.

O

occidentalis, Lachnus, 47.
oenotherae, Aphis, 118.
oregonensis, Aphis, 119.
oregonensis, Lachnus, 48.
orthocarpus, Macrosiphum, 66.

padi, Aphis, 94.

panicola, Schisoneura, 150.

pasaniae, Myzoeallis, 26.

pastinacae, Siphocoryne (Aphis, Ey-

adaphis), 133.
pastinacae, Hyadaphis, 132.
Pemphigus, 141.

alnifoliae, 146.

balsamiferae, 142.

betae, 142.

calif ornicus, 144.

coweni, 13.

fraxini-dipetalae, 146.

populicaulis, 143.

populiconduplifolius, 145.

populimonilis, 145.

populi-transvcrsus, 143.

populi-transversus, 143.

radiciooia, 141.

ranunculi, 144.

venafuscus, 146.

Pentalonia, 78.

nigronervosa, 78.
persicae, Rhopalosiphum (Aphis,

Myzus), 179, 185.
persicae-niger, Aphis, 119.
Phorodon, 79.

canZmmwn, 73.

galeopsidis, 81.

humuli, 79.

scrophulariae, 80.
Phyllaphis, 12.

coweni, 13.

fagi, 13, 178.

qucrci, 15.

quercicola, 15.
Phylloxera, 152.

popularia, 153.

salicola, 153.

vastatrix, 152.

vitifoliae, 152.
Phylloxerina, 153.

popularia, 153.

salicola, 153.

pinicorticis, Chermes, 152.
pini-radiatae, Lachnus, 48, 178.
pisi, Maerosiphum (Aphis, Nectaro-
phora), 66.
platanoides, Drepanosiphum (Aphis),

17.

pomi, Aphis, 109.
pomi, Aphis, 120, 179.
ponderosa, Lachnus, 48.
popularia, Phylloxerina (Phylloxera),

293.

populca, Pterocomma, 41.
populicaulis, Pemphigus, 143.
populicola, Thomasia (Chaitophorus),

36.

populiconduplifolius, Thecabius (Pem-
phigus), 145.

populifoliae, CJiaitophorus, 33.
populifoliae, Pterocomma (Aphis), 41.
populifolii, Aretaphis (Eichochaito-

phorus), 33.

populimonilis, Thecabius (Pemphi-
gus), 145.

populi-transversus, Pemphigus, 143.
populi-transversus, Pemphigus, 143.
Prociphilus, 146.

alnifoliae, 146.

fraxini-dipetalae, 146.

venafuscus, 146.
pruni, Aphis, 96.
prunifoliac, Aphis, 130, 179.
prunorum, Aphis, 121, 179.
pseudobrassicae, Aphis, 122, 179.
pseudotsugae, Lachnus, 48.
pteridis, Macrosiphum, 67.
Pterocomma, 40.

flocculosa, 40.

populea, 41.

populifoliae, 41.

smithiae, 43.

220

MISCELLANEOUS STUDIES

1 Hi iH-tat us, Myzocallis (Callipterus) ,

26.
pyricola, Eriosoma, 149.

querci, Phyllaphis, 15.
querci, Schizoneura, 15.
quercicola, Phyllaphis, 15.
quercus, Myzocallis (Aphis, Callip-
terus), 27.

B

radicicola, Trifidaphis (Pemphigus),

141.

ramona, Aphis, 122.
ranunculi, Pemphigus, 144.
rhamni, Aphis, 76.
rhamni, Myzus (Nectarophora), 76.
rhois, Rhopalosiphum, 86.
Rhopalosiphum, 80.

aohyrantis, 80.

arbuti, 84.

berberidis, 130.

corylinum, 81.

dianthi, 85.

hippophoaes, 81.

howardi, 86.

lactucae, 82.

nervatum, 84.

nymphaeae, 133.

persicae, 85, 179.

rhois, 86.

tulipae, 85.

violae, 86.
ribifolii, Myzus, 76.
ribes, Myzus, 75.

rosae, Macrosiphum (Aphis, Nectaro-
phora), 67.
rosarum, Myzaphis (Aphis, Myzus),

134.

rubi, Amphorophora (Aphis), 54.
rubicola, Nectarosiphum (Macro-
siphum) (Amphorophora), 77.
rubiphila, Aphis, 122.
rudbeckiae, Macrosiphum (Aphis), 67.
rudbeckiae var. madia, Macrosiphum,

68.

rufomaculata, Coloradoa (Aphis), 137.
rufulum, Melanoxantherium, 41.
rufulus, Cladobius, 41.
rumicis, Aphis, 101.

8

sabinianus, Lachiius, 49.

salicieola, Aphis, 123.

salicicola, Thomasia (Chaitophorus) ,
37.

salicicortieis, Symydobius, 39.

salieiradieis, Fullawaya, 35.

salicis, Siphocoryne, 132.

salicola, Phvlloxerina (Phylloxera),
153.

salicti, Cladobius (Melanoxan-
therium), 43.

sambucifoliae, Aphis, 123.
sanborni, Macrosiphum, 69.
Schizoneura, 148.

amerioana, 148.

lanigerum, 149.

panicola, 150.

querci, 15.

scrophulariae, Phorodon, 80.
senecio, Aphis, 123, 179.
setariae, Aphis, 124.
Siphocoryne, 131.

avenae, 84.

capreae, 132.

conii, 133.

foeniculi, 132.

nymphaeae, 133, 179.

pastinacae, 133.

salicis, 132.

xylostei, 133.
Siphonophora, 60.

acerifolii, 18.

ambrosiae, 60.

artcmisicola, 61.

chrysanthtmi, 62.

circumflexus, 74.

cucurbitae, 62.

heucherae, 64.

ludovicianae, 65.

solanifolii, 69.

sonchella, 70.

tulipae, 71.
smithiae, Ptorocomma (Chaitoplwrus) ,

43.

solanifolii, Macrosiphum (Siphono-
phora), 69, 179.

sonchella, Macrosiphum (Siphono-
phora) (Nectar opliora), 70.
sonchi, Macrosiphum, 60.
sorbi, Aphis, 108.
spiraecola, Aphis, 124.
spiraeella, Aphis, 125, 126.
stanleyi, Macrosiphum, 70.
Symydobius, 37.

agrifoliae, 38.

chrysolepis, 38.

macrostachyae, 38.

salicicorticis, 39.

tahoense, Cryptosiphum, 13.
taraxici, Macrosiphum (Aphis), 71.
taxifolia, Lachnus, 50.
tetrapteralis, Aphis, 125.
Thecabius, 144.

californicus, 144.

populiconduplifolius, 145.

populimonilis, 145.
Thomasia, 35.

crucis, 36.

negundinis, 36.

populicola, 36.

salicicola, 37.

viminalis, 34.

A SYNOPSIS OF THE APHIDIDAE

221

tiliae, Eucallipterus (Aphis, Callip-

terus), 21, 178.
tomentosus, Lachnus, 178.
Toxoptera, 129.

aurantiae, 129.

aurantii, 129, 179.
Trifidaphis, 141.

radicicola, 141.
Tuberolachnus, 45.

viminalis, 45.

tujafilinus, Lachnus (Lachneilla), 50.
tulipae, Macrosiphum (Siplionophora) ,

71.
tulipae, Ehopalosiphum, 85.

U

ulmieola, Colopha (Brysocrypta) , 148.
uhnifolii, Myzocallis (Callipterus),27.
umbellulariae, Hyadaphis, 133.

Vaeuna, 150.

dryophila, 150.

valerianae, Macrosiphum (Nectaro-
phora), 71.

vanduzei, Lachnus, 50.
varians, Myzus, 77.
vastatrix, Phylloxera, 152.
venafuscus, Prociphilus (Pemphigus),

146.

virburnicolens, Aphis, 126, 179.
viminalis, Aretaphis (Callipterus, Chai-

tophorus, Thomasia), 34.
viminalis, Tuberolachnus (Lachnus)

(Aphis), 45.
vinoae, Myzus, 74.
violae, Rhopalosiphum, 86.
vitifoliae, Phylloxera, 152.

W

woodworthi, Myzocallis, 27.

X

xylostei, Siphocoryne, 133.

yuccae, Aphis, 128.
yucoicola, Aphis, 128.

MUTATION IN MATTHIOLA

BY

HOWARD B. FROST

[University of California Publications in Agricultural Sciences, Vol. 2, No. 4, pp. 81-190]

MUTATION IN MATTHIOLA

BY

HOWARD B. FROST

CONTENTS

PAGE

Introduction 81

Genetic literature relating to Matthiola 84

Methods 85

Experimental data 89

The occurrence of apparent mutants 89

Characteristics and heredity of mutant types 92

1. The early type 92

2. The smooth-leaved type 118

3. The large-leaved type 125

4. The crenate-leaved type 127

5. The slender type 135

6. The narrow-leaved type 141

7. Miscellaneous aberrant types 143

8. Some probabilities of random sampling 145

General discussion 153

Summary 159

Literature cited .. .. 161

INTRODUCTION

It is hardly safe to use the term mutation without first defining it.
In this paper it will be taken to mean a genotypic change, or a change
in essential hereditary constitution, due neither to immediate cross
fertilization nor to segregation in a heterozygous parent. No attempt
will be made to restrict the term to any of the known or supposed
types of such genotypic change; a limitation of this kind, which
restricts the generally accepted sense of a widely used term, seems to
tend to confusion rather than to clearness.

fan

1'iM MISCELLANEOUS STUDIES

If we use the term factor mutation,2 (Babcock, 1918) where the
cytological change occurs within a locus, transforming a factor into a
different factor, two analogous terms will apply where the cytological
change is external to the locus. When the cytological change consists
of a loss, reduplication, or transposition of one or more loci it may
be called a locus mutation, and when the change consists in such
phenomena affecting a whole chromosome it may be called a chromo-
some mutation. If the term mutation is applied to the cytological
change itself, the last two types of mutation may be grouped together
as extralocus mutations, while the first type consists of intralocus
mutations. Examples of factor mutation are white eye in Drosoi>liild,
and probably the rubrinervis type in Oenothcra; an example of locus
mutation is (possibly) "deficiency" in Drosophila; and examples of
chromosome mutation are Oenothera gigas and 0. lata.

It is now evident that the immediate problem with Oenothera relates
to the mechanism of heredity in the genus. There are two sharply
opposed views. One is that recently emphasized by Atkinson (1917,
p. 254), when he says, "The evidence from Oenothera cultures points
more and more to the conclusion of Shull that 'a hereditary mechanism
must exist in Oenothera fundamentally different from that which dis-
tributes the Mendelian unit-characters.' ' The opposing view is
represented by Muller's (1918) strictly Mendelian explanation for
Oenothera, based on "an Oenothera-like case in Drosophila" ; he says.
"The striking parallel between the above behavior and that exhibited
in Oenothera makes it practically certain that this, too, is a complicated
case of balanced lethal factors."

A notable feature of the extensive genetic study of Oenothcra is
the lack of progress toward any definitely supported explanation of
its hereditary mechanism, which is not Mendelian. The only definite
non-Mendelian hypothesis of chromosome behavior so far proposed,
aside from "merogony" and other hypotheses (Goldschmidt, 1916)
apparently possible but not proved for Oenothera, which assume loss
of chromatin after fertilization, seems to be Swingle's (1911)
"zygotaxis, " proposed for the apparently parallel case of Citrus.
This suggestion that Ft hybrids may differ, apart from non-uniformity
of the P! gametes, because of the establishment of permanently differ-
ent arrangements of the chromosomes in the fertilized egg, still seems
to be purely speculative.

"With more or less "Oeiwth era-like" cases in other genera, the only
definite progress in analysis seems to have resulted from the assump-

2 Muller (1918) has recently used point mutation in the same sense.

f82]

MUTATION IN MATTHIOLA 225

tion of Mendelian segregation. With Oenothera itself, the trend of
the evidence tends to favor this form of explanation.

This fact is strikingly illustrated by two papers of de Vries (1918,
1919) which have appeared since the present paper was written,
especially as Muller's (1918) complete report on the beaded- wing
case in Drosophila (see especially pp. 471-474, 489, and 498-499)
indicates that de Vries had hardly yet realized the full possibilities of
the balanced- factor hypothesis. In the light of Muller's masterly
demonstration of these possibilities, we may be confident that "mass
mutation" is merely ordinary segregation, and that the "unisexual"
crosses of Oenothera are really "Mendelian" in their essential phe-
nomena. Some difference of usage respecting the inclusiveness of
the term Mendelian may be involved here, it is true, since apparently
de Vries would apply it only to cases where strictly homologous factors
are opposed in homologous chromosomes. Since, however (Muller,
1918), there are good reasons for expecting the occurrence of grada-
tions of similarity and of synaptic attraction between opposed loci, and
hence of gradations of linkage, the criterion of Mendelian behavior
should obviously be the occurrence of segregation between homologous
chromosomes, whatever their degree of similarity or amount of cross-
ing over. We have no reason to assume that an "unpaired" factor
in a parent would so divide as to be included in all gametes; on the
other hand, we have learned of a mechanism capable of insuring, in
certain particular cases, the inclusion of a certain factor or group of
factors either in every functional gamete or in every viable zygote.

No doubt, as Davis (1917) says, "A great forward step will be
taken in Oenothera genetics when types of proven purity have been
established . . . ." Meanwhile, cases of "Oenothera-like" heredity in
species known to possess the Mendelian mechanism deserve most
thorough investigation. Special interest consequently attaches to the
peculiar inheritance of certain apparent mutations of the ten-weeks
stock (Matthiola anmia Sweet), a species in which various character-
istics are typically Mendelian. A remarkable series of aberrant forms
in this species3 has been briefly discussed in two preliminary com-
munications (Frost, 1912 and 1916), and the present paper gives a
fuller account of the same phenomena.4

sin the variety "Snowflake, " a glabrous, double-producing form with white
flowers.

* While this paper was in press Blakeslee and Avery (1919), have reported
the occurrence of apparent mutations in Datura, which seem to be similar in
almost every respect to those here discussed.

F83]

226 MISCELLANEOUS STUDIES

Apparent mutants were first found in the course of work on
another problem, the relation of temperature to variation (Frost,
1911), conducted at Cornell University. Studied incidentally at
first, these new forms were later given special attention. About nine
thousand plants, of which about two thousand were progeny of
mutant-type parents of peculiar heredity (nearly one-fourth of the
latter representing crosses with Snowflake), have been examined
altogether. Some of these plants have been grown at Riverside, where
hybridization studies with mutant types are in progress. The present
account considers the origin and characteristics of these types, their
inheritance with self pollination, and the rather meager available data
relating to their behavior in crossing.

In connection with the work at Cornell, special acknowledgment
is due to the late Professor John Craig, and to Dr. H. J. Webber
and Dr. H. H. Love. Facilities for work were furnished by the depart-
ments of Horticulture and Plant Breeding of the New York State
College of Agriculture.

GENETIC LITERATURE RELATING TO MATTHIOLA

The work of Correns (1900) on Matthiola furnished one of the
earliest confirmations of Mendel's law, and also pointed to complica-
tions not found by Mendel. The earlier literature, according to Correns,
gives no indication of the study of Matthiola hybrids beyond the first
generation.

In his later paper on aberrant hybrid ratios, the same author
(1902) discusses complications in maize and in Matthiola. After
referring the deviations found in maize to selective pollination, he
considers a suggestion of de Vries relating to environmental modi-
fication of Mendelian ratios, and himself suggests the possibility of
selective elimination of gametes. He says (pp. 171-172), "Solche
Einfliisse brauchten nicht alle Sorten Keimzellen des Bastardes gleich-
massig zu treffen, sondern sie konnten eine Sorte starker angreifen als
die andere."

Von Tschermak (1904, 1912) lias made extensive studies of
Matthiola hybrids, considering mainly, as did Correns, pubescence and
flower color. The latter of these papers on hybrids in the genera
Matthiola, Pisum, and Phaseolus represents a careful analytical test
of the "factor hypothesis" of segregating inheritance, leading to the
conclusion that the applicability of this hypothesis is strongly con-

T841

MUTATION IN MATTRIOLA 227

firmed by the results secured. This work, with that of Miss Saunders.
leaves no possibility of doubt that the typical Mendelian mechanism is
present in Matthiola.

The most extensive genetic work on Matthiola is evidently that of
Miss Saunders, reported by herself (1911, 1911a, 1913, 1913a, 1915,
1916) and by Bateson and Saunders, with others (1902, 1905, 1906,
1908). This also is work on heredity in hybrids, with special emphasis
on the factorial interpretation of the various complications relating
to pubescence and to ' ' doubleness " of flowers.

Goldschmidt (1913) has explained the inheritance of doubleness
by sex linkage and lethal action of a femaleness factor in pollen
formation, and his interpretation has been criticized by Miss Saunders
(1913). I (Frost, 1915) have presented a somewhat different lethal-
factor scheme, and Miss Saunders (1916) has since restated her views
and criticized mine.

Muller (1917) has cited the inheritance of doubleness as a case of
"balanced factors," in apparent agreement with my formulation.

Apparently no one but the present writer (Frost, 1912, 1916; see
also review by Bartlett, 1917) has reported experimental evidence of
any notable tendency to apparent mutation in the genus, although
de Vries (1906, p. 338) mentions the occasional occurrence of vigorous,
rigidly upright individuals (a gig as type?), known at Erfurt as
"generals," and refers to the rare mutative occurrence of single
flowers on branches of double-flowering plants. Doubleness, and color
variations in considerable number, have evidently arisen under culti-
vation, probably through mutative changes.

METHODS

The general cultural methods employed for the first three genera-
tions have been very briefly described elsewhere (Frost, 1911).

The plants of the first four years were grown in pots in the green-
house. The plants of the first generation came from one or both of
two packets of commercial seed planted in the fall of 1906, and all
plants in the later cultures (possibly excepting series 18) were
descendants of these. The cultures will in general be designated by
the year in which the seed was sown ; the field and greenhouse cultures
of 1911 are indicated by 1911F and 1911 H respectively.

Part of the seed planted, especially in 1908, came from unguarded
flowers. The seed lots where this occurred will be indicated in the

[85]

828 MISCELLANEOUS STTIn

tabulation of parental data by italic figures, while protection possibly
defective will be indicated by an asterisk. It is not probable that
much vicinism occurred in the greenhouse cultures, since this plant
is well adapted to self fertilization.

In the first year's (1906) cultures the plants in each experimental
environment were separately numbered. Each plant was designalnl
by its number preceded by two letters indicative of the environment.
For greenhouse temperature these letters were C (cool), M (medium
temperature), and W (warm) ; for potting soil3 they were S (sand),
L (unfertilized "loam"), and G ("good" soil, fertilized). Thus CS1
CS2, WG9, etc., were pedigree numbers of the first generation, and
CG2-M8 and WG9-C10 of the second generation. A few syncotyle-
donous plants outside the regular cultures of 1907 were called WG9-
synl, etc.

For the work at Riverside a new system of numbering was adopted,
better suited to ordinary pedigree cultures, and the numbers from
this system are used below in the individual treatment of all but one
of the mutant types ("early"). This is essentially Webber's (1906.
p. 308) system, except that each initial or P^ individual of a series is
indicated by a letter; a full description has been published (Frost.
1917). With Matthiola each type or cross between two typos that is
tested receives a series number, the apparent mutants themselves
always being taken as the initial individuals of their selfed series.

The cultures of 1908 included progeny of various parents, one being
WG9-C10, an early and few-noded plant suspected of being a mutant.

The cultures of 1910 consisted of a second-generation test of WG9-
C10, and a first-generation test of other possible mutants, with control
lots. The plants were all grown on one bench in one greenhouse
(house C), from thirty lots of fifteen seeds each, lots 1—17 relating to
WG9-C10. The parents descended from WG9-C10 (see table 7) w.-n-
selected as those with fewest internodes, a medium number of inter-
nodes, and most6 internodes in each house of the 1908 cultures, earli-
ness of flowering being considered when parents were alike in number
of internodes. The control parents were both few-noded and many-
noded. relatively to their sibs.

In 1911 eighty progeny lots were grown in the field at Ithaca.
Lots 1 to 28, transplanted from the greenhouse, paralleled the test of

o Soil experimentally varied only in the 1906 cultures, temperature varied in
the two following years also.

« For house M, not the highest, which was exceptionally high, but the next
to the highest.

f86]

MV TAT ION IN MATTHIOLA 229

WG9-C10 made in 1910-11 ; all available progeny of WG9-C10, except
the crenate-leaved apparent mutant WG9-C10-C10, were tested, with
check lots between as before. Soil differences and unavoidable differ-
ences between lots in time of transplanting combined with hot weather
and drought to reduce the value of the results. The remaining fifty-
two lots, all field-sown, included a further test of the heredity of
aberrant types other than early. Most of these lots, however, were
progeny of Snowflake parents, grown to obtain evidence on the relation
of temperature to mutation and on the inheritance of doubleness of
flowers, and therefore the results are not reported here.

The 1911H cultures constituted a coldframe and greenhouse prog-
eny test of mutant types, mainly in the second generation, the plants
being grown in flats.

There was added in 1912-13 a small greenhouse test bearing on the
supposed mutative origin of WG9-C10, in view of the apparent possi-
bility that "WS1 or "WL10, in the same house with the unbagged WG9,
might have been heterozygous for the early type — cross pollination
then giving the apparent mutant.

Further progeny tests of the mutant types have been made in the
field at Riverside, beginning in the fall of 1913. Mainly on account
of the unsuitability of the usually hot and dry climate of River-
side, the cultures have been largely experimental and always on a
small scale, and germination or development has sometimes been un-
satisfactory. Cultures have been started in October, November,
January, and February, and a trial culture in progress at the time
of writing was started in August. Some of the plants of the 1915-16
cultures were kept until the summer of 1917, and many of them
flowered for the first time when about a year old.

In the cultures of 1913, growth was largely unsatisfactory, and
with part of the plants aphid injury interfered more or less with the
classification of types. In the cultures of 1914, the seeds were largely
lost through toxic effects favored by very shallow planting (as at
Ithaca) and strong evaporation from the soil. In subsequent planting,
the seeds, planted singly in small paper pots, were dropped into
relatively deep holes punched in the soil, and covered with sand.

The only field-grown plants closely resembling those grown in the
greenhouse at Ithaca, it may be noted, have been those of the 1917
cultures, grown in a lathhouse with added shade from muslin.

In the cultures of 1915-16, with partial shade and more frequent
irrigation than before, development was in general good; but even

[871

230

MISCELLANEOUS STUDIES

S 4)

t

"s

8 8 S

! j I I-H -H 4i

1 °
«

> 5

- 1
•8 s

1 e

a

§• 1
3 s

•*< 3

««a « .

& s

-u V
oj bo

5 as J
« ^ -

|

0

Si

V

!

ai«l A".iaA

i 1-1 :

ti 1 >-
efl No

g ^
- S 9

5 1 1

3
o
g

to aberran

.1 )i:| A'.I.I A
: H inns
pin: MII u.i

-uinu saAnai

r-t i— 1

A 8 *

s r^ *

> <D
^s If 02
0>

00 S

o

?>5

1

belonging

-siui sdi^-jtoq

i— 1 i-H

ft 0 g

- a .f c

»T B -g ®
T3 rt g oj

£.

Bi

|

O

g

£'

'H

_0

1

43.

i— i ': i— i i— i

1 ^ 1

S 5 ^ 2

- §, " &

e

rH §
H §
« t

|

Numbers

adA-j

|— «

* o 5

S s -i 1

v-' bo ^ 5

o o -^ S
rt S a *

El S
o

i

5^

sasa

i— i i i— i

^> CS *J _

fl ^ M o

Q> • ~T O • *4-l

to bo a/ <r ^^

0 q (H ^ W a

*3 1^^ S

a>

i-i

"s
g

p^SU,

V
1— 1 1-H I— 1 CO I-H C^

•^ rt ^ Q) oi _

C? "T? O -4^ rK ™

sig s,^ s

E J3 0 .2 -g .3

A

^

°6
<^j

paA^Us

CC : : C^ iO I-H i— i

CC ^s. fi
u <H J? 9 *>

x t« o o £ ,2
es _jj ^e ft r—

* .2 -S >> ^ £
a S * » t?

a

*H

<+-,
o

1

s

s|

11

a

*O 00 ^f l^- "^f OO O
OO CO CO I-H t>- O Tf
^H(M <N •*

«, « g i c -

^ o s 7r -5 s

55 „ ®° CL e
TO __5 35 « ^* ft

S-slI| I

5 Cj o tS -»J J3

3

o

C a -g
(-" *"^ ^

a o a o fl.2 2.2 c'-3
02 ca;

£ S 'w ^ o."^

^SBfc §1

^H^ 0 P „„

<— ( ^ o tiJD 4?
^ ^ "H ej c be
fcC * i 0 § as

1.8 - gS!H

5^ -«*_S *, t>

Parentage

> > -fa

|«5 o^ g,

•S § | o | 2.

• S * u " *
S

F881

MUTATION IN MATTHIOLA 231

here the mutant types, with one exception, often failed to grow satis-
factorily or to set seed. Infection, probably by Fusarium, evidently
was the cause of the death of many of these plants in their second
season.

With all cultures grown after (probably) those of 1906, special
care was taken to secure random samples of seed, and after 1908 no
plants were rejected. The only exception to this statement is the
rejection of one pot out of every fifteen, by number and systematically,
in the first twenty progeny lots of 1910. For the earlier cultures, a
certain amount of selection must be recorded, as follows. In 1906
the small and the largest plants were omitted at potting, and probably
any weak and abnormal seedlings had been omitted at the preceding
transplanting. In 1907 all markedly weak, late, or abnormal seedlings,
as determined mainly by the appearance of the cotyledons, were
omitted at the first transplanting; and the same was done in 1908,
except that certain lots from old seed were unselected.7

These last lots were arranged at transplanting in such a way that
the weak and abnormal plants came at the end in each lot.

EXPERIMENTAL DATA

THE OCCURRENCE OF APPARENT MUTANTS

In the cultures of 1906, 88 plants were grown to maturity, none
of these being suspected of mutation. In the cultures of 1907, among
170 plants one striking variant appeared; this plant, WG9-C10, was
exceptionally small and early in blooming.

In the cultures of 1908, 714 plants were available, including ap-
parent mutants in several hereditary lines as indicated in table 1. A
striking feature of the results is the scarcity of apparent mutants
among the seedlings classed as strictly normal at transplanting; prob-
ably the scarcity in the preceding years was due mainly to the rejection
of abnormal seedlings (see "Methods"). The first, second, and fourth
of these forms have been common in later cultures, while the third
and fifth have been rarer; the last three, if seen at all elsewhere, have
not being recognized as belonging to the same types as these three
plants.

7 One tiny plant from WG9, probably not viable, was discarded.

T89J

232 MISCELLANEOUS STUDIES

Table 2 shows the numbers and percentages of apparent mutants
found in the cultures of 1910 and 1911F. Since the early type seems
to differ from Snowflake only in size and earliness, and is probably
inherited without special complications, the available progeny of early-
type parents are included in the totals. The progeny of all parents
recognized as belonging to other aberrant types are omitted. The
second column under ' ' Percentage of mutants ' ' omits doubtful types
and individuals, but includes some individuals for which some doubt
was indicated in the original records. One rare type of 1911, large-

TABLE 2

Aberrant types: occurrence among progeny of Snow flake and early parents.
Apparent selective elimination at or after germination in
field-sown cultures."

Cultures

Progeny examined1'

Percentage of apparent mutants

All counted

Doubtful omitted

Greenhouse, 1910

338

5.03 =b .82C

4.14 =b .77

Field, 1911, seed

house-sown

2072

5.31 db .33

4.63 ± .31

All above

2410

5.27 ± .31

4.56 ± .29

Same,Snowflake par-

ents only

1364

4.33 ± .41

3.74 ± .38

Field,1911,seed field-

sown (parents all

Snowflake.)

3927

2.34 ± .24

1.55 ± .22

• Germination in greenhouse-sown lots, counting only plants examined for
type, 93.2 per cent; in field-sown lots, 45.1 per cent.

"Including some plants of uncertain type, indicated for some lots (when
apparently not Snowflake) in tables 1 and 3.

c For the calculation of these probable errors the percentages on the third
line are used as p.

leaved, here omitted, has proved to be genetic, but its determination
in these cultures was in general uncertain. A stricter criterion for
the second column, elimination of all individuals not considered posi-
tively determined, was used in the calculations for the tables for the
inheritance of the separate mutant types.

Evidently the more rigorous field conditions of 1911 eliminated
many of the "mutants'7 at or soon after germination. The "coefficient
of mutability" with good germination, as was the case with the un-
selected cultures of 1908, seems to be near 5 per cent, a surprisingly
high figure if immediate true mutation is responsible.

Before the aberrant types are considered separately, we may
examine (table 3) a detailed illustration of their occurrence in larger
cultures. It seems probable, from this evidence, that any descendant
of "WG9 was capable of producing any of the mutant types so far

[90]

MUTATION IN MATTHIOLA

233

1

1
rt- *

P »-

|1

co £
1
1
c

^>

3 T3

9 P

5 O
3 P

3 03

'S

!^
i*

Oi*>COtOt-'OcOOO^arf».COtOi-'OCOO5O>*>'COO5C»^IO5OiGC<ItOH-'

Field lot

TABLE 3
1911, field; plants transplanted from greenhouse. Aberrant types: occurrence among progeny of Snoivflake and early parents.*

^

o

CO

Generation 1

Ancestry

2 00

O CO Oi

Generation 2

^^^^^SSSggSQooo: ^^QS^^9S

i-i co ^r oo 01 co k-> 4». o> GO ^j to oi co ^ w OT ^ : to i— ^ w to i— £r "^

O H-> : ri^ O t*>- 00 ^

Generation 3

1— >

o
o

to
o
-J
to

O5i-'tO*.OitOOiOiOOO^I^JOOO5OOOOOCDCOOOCOOiOOOiVIO^J

Total progeny of
determinable type

H- '

g

to

~4

• )-* to H* • : h->h-' to to i— ' : to : tOCOCOt-1: i— ' i— ' i— ' h-> )-»

Smooth-leaved

Progeny evidently belonging to aberrant types

-u
co

co

: :h->:i— 'it—": : t— ' i— ' *—» : : : t— » : H^ i— ' : ::::

Small-smooth-
leaved

^1
^1

i— i

02

-4— to : : : : H-» >— ' : : to : i— ' : i— ' i— > : : i— ' i— > i— ' :

Crenate-leaved

1— »

o

^

Semi-crenate-
leaved

o

Oi

-

Pointed-crenate-
leaved

o

Oi

i—"

Medium-smooth-
leaved

CO
^J

to

o

: : : OitO: : : *—> to t— ' : .' H- » : tO H- ' i— ' : i— ' : i—i i— ' : : h- ' : :

Narrow-leaved

rf*

00

o

:•:::: to: ::::::: H-> h-- : : : : : CO H-> : to

Narrow-dark-
leaved

—
*•

CO

I—1 t— «

Slender

^: : : : : : ^^ : ::::::::::::::::::

—
0

to

h-" : : : ::::::::: i— >:::::::::

Small-convex-
leaved

1— >

CO

4*.

Compact

b

Cn

M

Curly-leaved

H- '

rf»-

co

. . ^. ...'-... •-*

Pointed-light-
leaved

• • ->s -*: : : "O

O

Oi

H- »

*-

Large-leaved

-*

^

CO

H-1 1— ' I—*

Medium-large-
leaved

-o • • — o -o •

h- >

*;

CO

to H-

Large-thick-
leaved

•^s -o • •

O

Oi

M

I—*

Small stout-
capsuled

^s

o

Oi

I— '

h^

Jagged-leaved

-c

Cn

CO

t— »

0

OOO4^(»C^tOOOiO5tOO5*-rf»-COhP>.tOH-'O500OiCOrfi.COtOOiOitOCO

All counted >

#>•

C3
CO

CO

c^

O5OCOOOOitOOCOOitO4>-CO4».tO4i-tOH-'Oi<IOiCOkti.COtOOi4».l-'CC

<<
Doubtful
omitted

[91]

234

MISCELLANEOUS STUDIES

discovered; the occurrence of the various types suggests a random
distribution among the progeny lots. This conclusion is confirmed,
and extended to CG2, by the field-sown lots of 1911.

Various parents belonging to mutant types have given other
mutant types among their progeny. There is some reason, as table 4
indicates, to suppose that parents of the early type have a more
marked tendency to produce these other types than have Snowflake
parents.8

TABLE 4

1910 and 1911F; sown in greenhouse. Apparent mutants among descendants

of WG9-C10 and other ancestors, comparing early parents (pure or

heterozygous) with SnowftaTce parents.

Ancestry

Type of parent

Progeny

Total
examined

Percentage of mutants

All counted

Doubtful omitted

WG9-C10

Pure Snowflake
Both
Both

i Early

( Snowflake
Snowflake
Snowflake
Both

1046

558
806
1364
2410

6.50 ± .47"

4.30 ± .64
4.34 ± .53
4.33 ± .41

5.27 ± .31

5.64 ± .44

3.41 ± .60
3.97 ± .50
3.74 ± .38
4.56 ± .29

a For the calculation of these probable errors the percentages on the last line
are used as p.

CHARACTERISTICS AND HEREDITY OF MUTANT TYPES
1. THE EAELY TYPE

So far as is known, "WG9-C10 (figs. 1, 2) was the only apparent
mutant of the early type in the cultures. Since, however, this
type visibly differs from Snowflake only or mainly in quantitative
characters, it cannot be positively identified without comparative
progeny tests, and therefore may have been represented by mutant
individuals not used as parents. WG9-C10 was much smaller pro-
portionately than were its progeny; this difference was probably due
to an embryonic abnormality, early blind termination of the main
axis, which was occasionally observed elsewhere and probably occurred
in this case. Plants of this type, as compared with Snowflake, are, in
general, fewer-noded, smaller, and earlier in blooming.

The principal data from the cultures of 1908 are shown in tables
5 and 6, which also indicate the later conclusions as to the segregation
of the early type in the cultures of this year ; figures 3 and 4 illustrate

8 Inspection of the data in detail indicates that this difference is not due to
the possible tendency in parents grown in the warm house to more frequent
apparent mutation.

F921

MUTATION IN MATTHIOLA

235

TABLE 5

Cultures of 1908. Time from sowing to emergence of corolla of earliest flower
of primary cluster. Frequency distributions.11

Parents:

Singles

Doubles

House C

House M

House W

House C

House M

House W

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

Day.- h

110

111

112
113
114
115*
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155

It
It

1 '

1

It

It

2
3
4
2
3
8
4
7
16
3
4
7
1
2
1
1
2
4
1
2
1
1
It

U

"it

1

1
2
2

2
3
7
2
5
7
4
3
8
7
7
6
4
2
1
3
3
3
1
2
2
3
2

It

1

1
1

5
7
9
8
8
12f
12
7
4
2
6

i

1

It

It

"it
It

1

i

2

3t
4
3
4
2

"U
Itt

i

1

2
4
1
7
10
18t
7
10
4
4

i

2

1
4
3
3
9
2
4
1
6
5
2
3

1

i

-2

i

3
4
8
13
9
15
10
9
6
5
4
1
1
2
1

1
1
1

1
1

i

4

2
2

1

1

1
1

1

2

1

1

1

1

1

• Daggers (t) indicate the position and number of apparent mutants. Double
daggers (t) indicate inheritance of parental type (here, early); all single progeny
of WG9-C10 here reported have been tested for inheritance of this type. The
conventional statistical constants corresponding to the house totals of tables 5
and 6 have been published (Frost, 1911); the means for flowering given there
are too high by one half-day.

b To time of observation (upper limit of one-day class).

F931

236

MISCELLANEOUS STUDIES

TABLE 5. CULTURES OF 1908 (Continued)

Parents:

Singles

Doubles

House C

House M

House W

House C

House M

House W

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

Days b

156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173

1

It

It

• Daggers (f) indicate the position and number of apparent mutants. Double
daggers ( + ) indicate inheritance of parental type (here, early); all single progeny
of WG9-C10 here reported have been tested for inheritance of this type. The
conventional statistical constants corresponding to the house totals of tables 5
and 6 have been published (Frost, 1911) ; the means for flowering given there
are too high by one half-day.

b To time of observation (upper limit of one-day class).

the difference in earliness between the early and Snowflake types. The
parents grouped under "rest" include CG2 and WG9 themselves, with
four progeny of the former and eight of the latter. Of these fourteen
parents, not one has produced exceptionally few-noded progeny like
those of WG9-C10.

Apparently WG9-C10 was heterozygous for a " f ew-nodedness "
factor not carried by any of the other parents tested. Neither in
the 1907 cultures nor in the 1908 cultures now under consideration
did the data suggest that WG9 itself was similarly heterozygous.
Tables 5 and 6 include the first 30 progeny of WG9, for each house,
as arranged at the first transplanting,9 88 plants altogether ; including
the remaining plants, mainly weak or abnormal at transplanting, the
total is 116. One of the F, plants (WG9-syn3-M10) was very sug-
gestive of the early type, but (tables 12 and 13) it gave only Snow-
flake progeny in a small test.

9 See page 89. Two plants not producing a normal main inflorescence are
omitted.

F941

MUTATION IN MATTHIOLA

237

TABLE 6

Cultures of 1908. Number of main-stem internodes below first flower-bearing
node. Frequency distributions.*

Parents

Singles

Doubles

House C

House M

House W

House C

House M

House W

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

WG9-
C10

Rest

Internode.

16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62

It

m
it

1

m
it

1

1
1

9
25
29

i

2f

it

1

2

1

2t
9
14
22
27
5
3
4

i

i
i

5
6
8
5
13
6
8
6
12
9
6
3
3

i

7
17
24
13
8
2
1

2*

1

1

22
6

i

1
1
2
7
15
16
19
' 4
1
8
3
1

it

1

2

It

It
It

It

1

'it

i

It

It

It

4
3
4
3
6
3
10
6
2
3

8
1
1
3

1

1

1

at

2

1

1

1

1

*• See note a to table 5.

f951

238 MISCELLANEOUS STUDIES

The differentiation of the early race is very marked; with the
singles, in fact, the later cultures indicate no case of overlapping in
the 1908 cultures, in either character, between extracted pure Snow-
flake and pure or heteroz3rgous early. The total sterility of the doubles
necessarily leaves their constitution somewhat in doubt.

The cultures of 1908 so far suggest that WG9-C10 was a mutant.
To be reasonably certain, however, we must have further evidence
(1) on the fact and mode of inheritance of the supposed new type,
and (2) on the possibility that either WG9 or some other plant of
the cultures of 1906 brought the character into the cultures. We shall
now consider somewhat extensive evidence bearing on these points,
concluding with a special test of the possibility of vicinism.

When I last saw the warm-house plants of 1906, three were known
to be singles, and all but two of the rest were recorded as certainly
or probably doubles. Seed was secured from these three singles only,
and presumably no other singles occurred in the house. Since this
seed was all from unguarded flowers, we must consider the possibility
that WS1 or WL10, the other warm-house singles, brought the early
factor into the cultures. It is also barely possible that pollen was
brought to WG9 from some plant not in this greenhouse.

These two parents were tested in supplementary cultures, in house
C in 1907, and in house W in 1908. The 1907 progeny averaged
slightly earlier than those of other parents, but this may have been
due to their position, which was much nearer a partition separating
the house10 from a warm greenhouse. Unfortunately the internodes
were not recorded.

In the 1908 cultures these lots were potted two days later than
most of the other lots and one day later than the WG9 lot, and for
some unknown reason the WL10 lot wilted badly for some days. The
parents in question gave singles (16 and 11 plants respectively) which
when compared with progeny of CG2 and WG9 (23 and 15) might
suggest that the parents were heterozygous for the early type. The
results with the similar numbers of doubles decidedly disagree with
these, and suggest that cultural accidents produced the differences;
the WS1 lot was not exceptional, while all the WL10 progeny were
grouped near the lower end of the range of the other lots. In view of
all the facts, the data hardly deserve tabular presentation, but they
raise a question requiring further study ; a later test is reported below.

10 A temporary substitute for the regular house C.

[96]

MUTATION IN MATTHIOLA

239

,1

I HS
d cr 2

OR5 »• 3

00 2 ^

V)

t±o
^9=

P co

-

d o

00^

f
o

^ £2 o 9 .0

o *° °

fill I

0
<o
O
»

1 i. i a I

lilt 1

ration 1

A 1 * i

§rtottbh-i'K*ts5oto£-> §3 ^^-s o o o

o

1— »

/^

thnsi I??? ffisfffgi1

* M i a

(Snowflak

3 eneration

g

• p^-2->2' -£- ST' (5 -2^ — (6 (B fD

(6

to

3 ^^ ^ N~^

68

p ^_^

i

- ® H- O *^ ' O

60 CO CO

2 ^ »? 3 cocc^sgccoiaitTjMp^^^HM^220^^

H"S^

O

m

I'll o 1 1 i ^ isl 1 1 ^i i 1 ^^1 1 i<T

CLo o

a
•-j

C^" ^J. i-^-i J^ |^ j-y^ ^" ^) ^^ ^ i i,^^ ^^ ^+^ |~j^ j^^S ^S J^ 2. f^^ «*

•v ^ j^j ^|

to

|l| fft^^ffl Slf Iff

^^
-JL^

o'

CO

(5 ^ <6

£e-

— '*—*-'

rf». 4^. tOh^tOCOCO CO Cn tO CO *» tO tO CO tO
00 t£> tO^It-'OO CO rfi-VI *. tO 00 M O5 *-t

CO

to

13

i3

Oi I— ' GO Oi rf» tO 4^ O C5 tO O O Cn CO O .—

h^

S ?"

3

5"

E

)__A_% ^>~- ^ ^ _x . ^J^-,. «_

^r-*—,

3'

SJ"

31

^
"^
p

•<! COCOtOrfxi-'rfs.rfi.COtO.^jCTitOCOCOCn^COtOtOCOtOtOCOCOtsStO

H-iCOCO

< »
SIS

1

CO rf»-OOCnCOOl— 'tOCOOl tf^-<lCOOOCntOCOCOMrf^Cr<i— 'OCOMO

OiOCn

gs

§

EL

cp

21

d

CO CO M CO O5 CO rf>. -<| O5 ~4 <l Oi 4^ <l OS OKI QD Cn CO CO O5 OS *». <l <l GO

coco-^

0

S
cr

5"

a>
Z

O

o
«

H- 00 0 tO CO O M <l<l O5 O5 rf»- O Cn 00 00 VI t^. CO Oi O CO 00 O <I ^1 OS

O5 Ol O5

o
c:

i

2

9

y

1

o-

[97]

240

M I S< ' A 1. 1. A \ KOUS STUDIES

In the cultures of 1910 and 1911F, all the 1908 progeny of A\7G9-
C10 were tested. On account of the variable nature of the quanti-
tative character involved, an elaborate study was necessary. Only
small cultures could be grown in the greenhouse; these were supple-
mented by larger lots in the field in 1911, but inhibition of flowering
by the hot summer, together with the effects of disease and soil varia-
tions, made the field results erratic and necessitated special methods
of treatment of the evidence.

5.">
50
45
40
35
30
28

in

16

10
5

Singles — —
Doubles _ _ _

•

•

•

•

•

•

•

•

•

•

•

_-_

M3 M4 C2 Co Cl

1 1 1 1 1

M6 M9
C5

M4 M2 M7 C3 C7 VV6 Wo W10 M5 MS

1 1 1 1 1 1 1 1 1 1

Cl CIO

1

1 1
CIO C9 CIO C9

1 1

WG9

Ancestry

Chart 1. Cultures of 1910. Internodes: parental values and progeny means
(respectively shown by dots and lines) for progeny lots 1 to 17, omitting
aberrant progeny. Parental values should be compared only for the same house.

Table 7 gives the available data for the parents of the 1910 cultures,
and the numbers of progeny available for quantitative data. The
order of the pedigree numbers here is the same as that of the progeny
lots on the greenhouse bench. For convenience, the 1910 tests of other
mutant types, together with tests of several Snowflake parents, are
included in the table (lots 18 to 30).

F98]

MUTATION IN MATTHIOLA 241

The plants were grown in house C of the previous work. Two or
three plants (one shown in fig. 25) were extremely vigorous, pre-
sumably because of some accidental soil difference; aside from these,
a few apparent mutants, and a few plants otherwise abnormal, the
plants were fairly uniform except where heterozygosis was to be
expected.

The data for time of flowering, as with the 1908 cultures, show the
same main features as the internode data, and only the latter will be
considered in detail. The types were again more widely different in
internodes than in earliness, a fact which seems to indicate that the
early type grows more slowly than Snowflake.

So large and so regular are the differences in internodes that the
means of these very small lots seem worthy of presentation (chart I).11
Apparently the few-noded character was carried, among the nine
parents descended from WG9-C10, by all except the three parents
having the highest numbers in their respective houses.

Tables 8 and 9 give the internode frequencies for the singles and
doubles respectively, by separate progeny lots and by groups of similar
ancestry. The range of variation for the check lots, omitting the
indicated apparent mutants and other apparently abnormal plants, is
rather surprisingly small, as is the case with the cool-house cultures
of 1908. The three late progeny of WG9-C10 give lots closely corre-
sponding in range to the check lots, only one individual falling below
the range of the combined check lots. The six early and medium
progeny of WG9-C10, on the other hand, give distributions of far
greater range than do the check parents, extending to much lower
values.

Tables 10 and 11 give the ordinary statistical constants for the
grouped lots. The mean number of internodes, for both singles and
doubles, is about 25 per cent lower in the progeny of the six few-
noded parents, the difference being not far from ten times as great as
its probable error. The increase in variability with the progeny of
the early parents is also striking, and the difference is about five to six
times its probable error. With time to flowering, it may be noted, the
differences are similar to those with internodes, but somewhat less
marked in the case of the mean ; the flowering data are not given here.

It is plain that the previous conclusion as to the heterozygous nature
of WG9-C10 is sustained. The elimination of the apparent mutants

n Calculated with the apparent mutants and four other apparently abnormal
plants eliminated; see tables 8 and 9.

T991

242

MISCELLANEOUS STUDIES

S

rO

1

e

'C
O

«9>

1

O

g <R.
S §

O-5

ug

.-iCOiOO^TjtcOCO^t-lOJCOlOCOi-i-^lOINCO

: CO •* O» as 1» CO —i : «-i

^H CO »0 O •* ,H (N (N CO -H <N -H — i CO ^H (N

ot^oocs O ^

"S -^ no .
-*

^ 5 60

u ~ O

p | 03 '"^

§ s >

sow

a ° • o
c J <w _w ao

w 2

.

-• _

^

.

> S
"2 o o

lS .2° §
S ^=03
»- H I o

^

[100]

243

ce
«c

J

o

— "°

<j

t-< C<l C^r-tj-ti-l i— 1 C^

O • -4J

05 fJ tn
E T3T!

O5

w-o

•

1

b

Ofe

1

1

1

1 ^H 1-1 ,-H T-l T-H rt |

0 _ 'g

C

o

•^H- ^- : : : : i-

- J

s

M

&

'

•S

0

S *^ 0

rO

O

3

•a $3

-^

O

o-2

^_

- -1-

•2

'S

;z-

:::::::::::::: I-H

1 !"*

|>

o

o

« 5 .S
?3 "S

§

O

0

•~ 9 rQ

* g 01

&

00

03 3
qj ©

1

8

^

« 2 %

10

.2 'S •£

rrt SD

^
ts

%

•S » ^o

0

s

o

1

is

^ T3 *

•£

0

«

.O Q^ -*J

«

A

O

1*

!g3§

L

to

5 ? a

2 ** _S >».

i*

ft.

^

:•»-• :::::::::::::

-*J ^5 i 6C
- cj o

e

o

•^

t. -C —

ss.

O)

O

rt "^ *w
pij O O3

3
o

£

PnS . -"S

C8 O P^

«

rO

^

on
V>

X

^ O S

i_ as t* ^^H

^S
O

o

N

g

0

5

s a ^-^^

tw>

s

Tf

c ® S3^^

s£

a

'a * ^ cs ro

6

05

?

%

«J oil

•<?»

O

(0

S

^

o^ * 1 *

0

,_,

p< pj ^®

k

6

O

^3 as o ^ °°
•u 5° •— <

S

o

ft

^_

* o fc

3

fe

O

0

:::::::::::::: i— i

C8 'O rQ -H

IN

<»

O

.S | -Sj-S

o

0>

. . .^_

2 .PH 9 •& PL,

S

::: i— i :::::::::::

S).S §0

£

O

co

^H ^_

^Ig "go S

o>

«

^H ::::::::::::: 1-1

1 g

<sT

0}

H

a

V

O

IN

O

CO

d
v
O

||c?^^c?c?5^^^^^5g^^^

J3 g 3 O "g

H S HI S
« ° « ^

"S £ fc

etf P- -iJ

[101]

244

MISCELLANEOUS STUDIES

-o

1-1 CO CO •* »O M •-< ~H -^ 1-1 CO Tf »-i C^ 1-1 1-*

•< .2

: C^ 00 1-1 : 1-1 1-1

i<N<NC^J<N(N(MC^<N<M

Oi O 1— *

<N COCO

BE

CO 00 CO OO CO 00 OO CO CO

[1021

MUTATION IN MATTEIOLA

245

Doubles

Singles

^

A

r "^

"' 'A

>> ^

>>

^

0

^^ O

^- <—•

O

p

Kl
^O

"j^"^1 ^

'^•'"^

CO

~S-o

•ib-S

**

j5

«

3

a

C6

9

Z?

>< ^A_

>*<

_^^

0 |

1

1 %$

o'

g

gr

(O £

<^

*
1

><^

><^

o

|i

~'§.Mtrlt7l

""^ 2

H t^tr1

p

o a

H-Srl

o'

g

i-< r*- <r>-
t^T ^ ^

at

-I, I

f

c*-

S. ^

CD
CO

S

^ §.'

o* ^

ll

1 1

to

1— '

t-«cooo to en

00 *-4

^ H-> Cn

O D

-^

4^ tO

Co Cn ^^

(D ft

O QJ

3 n

Oj «>

a

ll

to to to to to

Cn Cn O OS Os

to to

to to to

to co co

I— > --j co i<^ en

co rf».

Cn O 00

s g4

CnOO tO*-O

OS O O

**

Co ^

a

ns* cS

H- If If H- If

H- If

If If If

v>

s ^

•— to ^ to to

00 tO CO COM

toco

OS tO

OS *-00

»— to to

II.

3 <§

s" °

CO &4

CO CO 4x tO 10

Cn 4-

Cn to CO

If?

CO tO Cn H- CD

i— to cn rf*. Cn

too

101-

co oo *••

H-CDM

a. as
2 P

Oi

If If H- If H-

H- If

H- H- If

S'o.

S

ll

§

t-i H- ' 00 tO 1— >

co oscn Oco

I— tO

00 CO

OOCD tO

So"

0

--«>

S1

«.

i-» h-> tO I— '

h- ' (-»

to i-«

Cn to H- 00 i—

0000

OSOOh-"

O

^

gs^gs

o2

to to bs

I-* OS OO

11

«.
1

JS'2.

H- H- If If If

H- H-

If If If

gl

1

h- >

tOH-

o

in OS M M M

OSM

O O •*!

*^*

to to ^ oo oo

0000

00 tO^l

[103]

246

MISCELLANEOUS STUDIES

C^ W ^O &3 O C5

COO CO -^ QCCD

0

a

^-»^H ^-(

$

|

n ti ij n n ij
n n 11 "n n n

i

c?

«

rocor- co"3W

o

>

O* CN x

0

Tf OOO CO 00 O

(N IM

00

O O5 >O i-l ^H CO
O C*l CO »~H C^ CO

3

"H

a

1

«

•tt

'%

•H -H -H -H -H -H

rQ

K

0

COCO lO CO I-H (N

*>

•S

r-l C<l •* i-H W Ttl

§

£>

§

cogo coo^.

1

B
1

•H -H -H -H -H -H

3

(N O b- C<l O CO

G3 O O3 O ' * ' "

berran\

00 CJi-l CO CO O
IN(N(N (N(N (N

s

3

i

1

Oi *O 00 CO ^^ CO
^ ^H CO *O C^ CO

I

n

ss. SS£

!•

a

^3w 3^w

e

O

_ -

-c

§

b

— — — —

tt

§

d

9

ijo >2o

§

B

V—Y—/ —V—'

»a

ad

"5

-u

O

-*

O5 O5

CD

EJ

o o

a>

a

"o

^* ^

c

«

M

sa[8uig sajqnoQ

U04]

MUTATION IN MATTHIOLA 247

and the other abnormal plants presumably gives a better comparison
as to mean and variability, but the conclusion is the same in either
case. The three many-noded (late) parents descended from. WG9-
C10 give no definite indication of being genetically different from the
"check" lots not descended from WG9-C10, while the variability con-
stants are sufficient, taken alone, to make probable the genetic differ-
entiation of the fewer-noded progeny of WG9-C10. Apparently all
the fewer-noded progeny of WG9-C10 that were tested — seven, when
WG9-C10-C10, a crenate-leaved apparent mutant (tables 12 and 13),
is included — were either simplex or duplex for presence of an earliness
factor or factors.

The variability of all the thirty progeny lots, taken together, is
high, as might be expected, though decidedly below that of the progeny
of early parents. This high variability is due only in very small part
to the progeny of the five or six supposedly mutant parents; the last
thirteen lots, alone, are much less variable than the mixed early lots.
The portion of the cultures containing these progeny lots from
aberrant parents was conspicuous for irregularity of germination, and,
on the whole, a relatively low rate of germination.

A few of the last thirteen lots give more evidence bearing on the
origin of WG9-C10. The early WG9-syn3-M10 (tables 12 and 13)
gives no evidence of genotypic differentiation from its ordinary sib,
WG9-syn3-Mll ; WS1-W216, another phenotypically early parent,
also failed to transmit earliness 'to its progeny. CG2-C2-C6, on the
other hand, although itself an ordinary plant, shows a rather sus-
picious tendency to the production of early and few-noded progeny,
but better evidence would be required for any positive conclusion.
WG9-C10-C10 appears, from the data in tables 12 and 13 and from
observation of the flowering of plants of the next generation in the
1911H cultures, to have been heterozygous for the early type, as well
as for the crenate-leaved type. We find in this test no definite indi-
cation that the early type has appeared elsewhere than in WG9-C10
and its descendants.

The F2 progeny of WG9-C1, an abnormal plant whose Fx progeny
were unusually and uniformly early but not few-noded, have been
included with the other check lots without question. This treatment
seems justified by the flowering data, which do not indicate any
repetition of the precocious development of the first-generation plants ;
the peculiarities of the Fl cultures, if not a mere cultural accident,
presumably depended on the very abnormal development of the parent,

fl05]

248

•MISCELLANEOl .S ,s777>/A',s

TABLE 12

1910, greenhouse, lots 18 to 30. Number of main-stem internodes below first
flower-bearing node. Frequency distributions for singles.*

Gen. 1

CG2

WG9

wsi

WL10

Gen. 2

C2

W4

syn (M) 3

C228

W2

CIO

W2

W7

W216

W?25

W220

Gen. 3

Inter-
nodes

21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44

C6

W3

C17

M10

Mil

M7

CIO

W2

C5

1

It

1+

It

1

i

i

1

1 1 1 1 1 1 1

2tt
It

1

"2"

3
1

2
4

"l"

2

2
3
3

2

1
1
3

1
2
2

1
1

1
1

i

"it

1

1

1

1

1

It

1

1

1

1

It

1

It

It

• See table 7 and notes to tables 5 and 8.

TABLE 13
Same as table 12, for doubles."

Gen. 1

C

32

• W

G9

W

SI

WL10

Gen. 2

C2

W4

s

yn (M)

3

C228

W2

CIO

W2

W7

W216

W225

W220

Gen 3

C6

W3

C17

M10

Mil

M7

CIO

W2

C5

Inter-
nodes

16

1

17

2

18

19

1

20

21

3

22

1

2

2

23

1

2

It

3ft

1

24

2

1

It

If

2

25

4

1

3tt

2

It

If

1

1

1

26

2

2

2

1

2

If

27

2

1

2

2

3

3

1

2

28

1

1

1

2

5

1

29

2

1

1

2

30

3tt

31

32

1

1

It

33

34

1

1

Ances-
try

• See table 7 and notes to tables 5 and 8.

[106]

249

with its aborted main axis and very late production of a flowering
shoot.

Table 14 shows the general plan of the house-sown field cultures
of 1911. The progeny of "WG9-C10 were arranged as before in the
order of their numbers of internodes for each house of the 1908 cul-
tures, beginning with the lowest numbers. The parental values for
flowering and internodes are the values indicated by "$" in tables 5

TABLE 14

1911; field, plants transplanted from greenhouse. Ancestry, seed, and
numbers of progeny*

Lot

Ancestry

Seeds
sown

Number of
plants alive
33 days after
sowing

Numbers of plants for data
on mutation and flowering

Gen. 1

Gen. 2

Gen. 3

Totalb

Singles

Doubles

11

fier

/C8

80

79

77

34

43

2/

CyO

\ CIO

71

71

70

36

34

• ]

fC2

80

80

79

39

39

4 [
5f

CIO

j C5

1 C8

80
80

79
79

78
78

40
35

38 .
43

6J

LCI

80

78

76

36

40

7\

/"*ic

/W18

80

79

77

37

40

8J

C5

\W24

80

76

75

41

34

91

'M4

80,88

77

76

30

45

10

M9

80

80

78

36

41

11

CIO

M6

80,63

77

77

34

42

12 k

M2

8Q,71

78

77

36

40

13

M7

80

80

80

37

42

14 J

M8

80

74

70

33

37

WG9

15 \

/T»

/C3

80

78

78

30

48

16 /

uy

\C7

80

76

75

32

43

17]

'W6

80

74

70

31

39

18

W4

80, M

70

65

26

38

19

Wll

80

78

76

34

42

20

W9

80, 19

76

72

24

47

21

CIO

W5

80

79

75

33

42

22

W8

80, 14

76

74

36

36

23

W7

80

75

72

32

40

24

W3

80

73

71

37

34

25 J

W10

74

72

66

32

34

26

CIO

80

60

59

27

32

27 \

/~Tk

/ W10

80

74

73

33

39

28 /

uy

\W24

80

80

78

32

45

a For plan of arrangement and parental data, see page 86 and tables 5 and 6.
Seeds from unguarded flowers are indicated by italic figures; where two numbers
are given the first is the total.

b Including twelve plants (all late mutants) with which determination of the
form of flower was impossible.

[107]

250

MISCELLANEOUS STUDIES

and 6, in the order there given, except that the arrangement by inter-
nodes reverses the two-day difference in earliness of the parents of
lots 19 and 20; for convenient comparison, the parental and parent-
lot internode values are included in table 19.

Two progeny lots were set in each of the fourteen rows; probably
the soil was less favorable at the east end of the plot, and hence for
the even-numbered lots, at least in about the last seven rows out of the
fourteen.

The plants were beginning to grow very rapidly when moved to the
field. On account of deficient soil moisture and excessive heat, the
transplanting was slow and in part purposely delayed, covering a
period of five days. Lots 21 to 28 were set three days later than lots

80
3 70
I 60

k 50
Si 40
g 30

V

* 20
10

ODD

— —

"---•

.

._...

.....

-—

1 2 3 4 5 6 7 8 9 10 11 12 13 14

(C) (C) (C) (C)
Row number

Chart 2. 1911, field; lots transplanted from greenhouse. Percentages of
progeny lots not flowering by November 3, for singles. Apparent mutants and
injured plants eliminated. Odd-numbered lots represented by solid line. (C)
indicate check rows. The curves are broken between rows 10 and 11, where a
cultural difference enters.

11 to 20, and the later loss of roots resulting seems, in spite of rain
coming the next day, to have seriously delayed flowering. Lots 1 and 2
wilted badly after transplanting, and some difference in soil con-
ditions in the flats, rather than a genetic difference, was doubtless
responsible for the exceptional lateness of these lots. Lot 20 lost an
exceptionally large leaf area as a result of transplanting. A fungus
disease (a slow stem rot) was more common on lots 20 to 24 than
elsewhere; it doubtless killed some young plants and delayed or pre-
vented flowering in some other cases. Possibly the soil was poorer in
the later rows.

[108]

MUTATION IN MATTE 10 LA

251

TABLE 15

1911, -field; plants transplanted from greenhouse. Plants alive November 3,
not having flowered. Singles.

Row

Lot

Non-
flowering
plants

Non-flowering,
Snowflake and
early types*

Lot

Non-flowering
plants

Non-flowering,
Snowflake and
early types8

1

1

27

26

2

29

28 (27)

2

3

5

2

4

7

5

3

5

9

9

6

11

11

4

7

7

5

8

17

17

5

9

0

0

10

2

2(1)

6

11

1

0

12

9

7

7

13

19

18(17)

14

20

19

8

15

12

12

16

14

14

9

17

1

1

18

8(7?)

7(6?)

10

19

0

0

20

3

3

11

21

8

8

22"

11

10

12

23

21

20

24

23

23

13

25

16

14(12)

26

14

12

14

27

23

22

28

24

24

a Omitting non-flowering apparent mutants. For the numbers in parenthesis,
"doubtful mutants" are classed as mutants. Two plants accidentally seriously
injured, in lots 14 and 25, were counted out with the mutants.

" The stem-rot disease (see p. 108) was evidently worst in lot 22; some two or
three of the worst infected plants (included above) were nearly or quite dead
by November 3.

TABLE 16
Same as table 15, for doubles.

Row

Lot

Non-
flowering
plants

Non-flowering,
Snowflake and
early types*

Lot

Non-flowering
plants

Non-flowering,
Snowflake and
early types*

1

1

4

2

2

0

0

2

3

3

1

4

1

1(0)

3

5

3

3

6

0

0

4

7

3

2(1)

8

3

1

5

9

2

0

10

0

0

6

11

0

0

12

0

0

7

13

4

2(1)

14

1

1

8

15

2

2

16

6

5

9

17

1

0

18

3

3(2)

10

19

0

0

20

2

2

11

21

3

1

22

1

1

12

23

5

5

24

4

4

13

25

1

1

26

7

5

14

27

1

1

28

10

8

a See notes to table 15.

[109]

252

MISCELLANEOUS STUDIES

Altogether, these cultures are doubtless much less reliable for their
size than the greenhouse tests of the early type, but they nevertheless,
with due consideration of the points just mentioned, seem to permit
of fairly safe conclusions for most of the parents.

The plants were examined for flowering every other afternoon from
July 4 to November 3, inclusive (73 to 195 days from sowing) . A very
large part of the plants flowered in July, some in August, and a few
still later. Evidently the high summer temperature largely inhibited
flowering ; many of the singles and a few of the doubles entirely failed
to flower.

100

00

so

70

50

40

30

10

Chart 3. 1911, field; lots transplanted from greenhouse. Percentages of
progeny lots with primary cluster flowering or aborted by October 10-16, for
singles. Lines as in chart 2.

Figures 5 and 6 show the plants in July. Growth was usually
vigorous through the season, but the internodes were very short, the
branches numerous, and the region of the terminal inflorescence often
abortive, so that determination of the number of main-stem internodes
was not practicable. The emergence of the earliest corolla on the plant
was recorded at the bi-diurnal observations, and at two periods during
the season the aborted primary clusters were noted.

The data show very definitely the transmission of "earliness" by
the fewer-noded progeny of WG9-C10. Tables 15 and 16 show the
numbers of plants alive, without having flowered, on November 3 ; the
figures are thus a measure of lateness. The two progeny lots in each
row are given one line in each of the tables, in order to facilitate
separate comparison of the fourteen lots in each end half of the plot.

£V£

V

ODD

1

(C)

L™.

2 3 4 5 6 7 8 9 10 11 12 13 14
(C) (C) (C)
Row number

[110]

MU1ATION IN MATTHIOLA

253

The last column, with the apparent mutants omitted, no doubt gives
the best comparison. The data for the singles, reduced to percentages,
are also given in chart 2.

The doubles, which are often earlier to flower than the singles under
unfavorable climatic conditions, flowered so generally that table 16
presents no significant differences. The singles (table 15), however,
give definite evidence of segregation ; the lots in rows 2, 5, 6, and 9 to
11 all show a tendency to early flowering. Lot 26, consisting of Fa

TABLE 17

1911, -field; plants transplanted from greenhouse. Singles with primary
inflorescence flowering or aborted as indicated.*

Row

Lot

Aborted by
July 29

Flowering or
aborted by
Oct. 10-16

Lot

Aborted by
July 29

Flowering or
aborted by
Oct. 10-16

1

1

0

0

2

0

0(2)

2

3

12 (13)

19 (22)

4

20

24 (26)

3

5

2

2

6

4

8(9)

4

7

2(3)

3(4)

8

1

1(2)

5

9

22

27

10

26

33

6

11

25

29 (30)

12

17

22

7

13

1

2(4)

14

2(3)

2(3)

8

15

4

5(6)

16

0(1)

1(2)

9

17

19 (20)

20 (23)

18

9

12

10

19

25 (27)

29 (31)

20

11

12

11

21

4

6(8)

22

7

9

12

23

2

3(4)

24

1

1(2)

13

25

0

3(4)

26

3

4(5)

14

27

1

3(4)

28

0

0(1)

a In this table and also in table 18 the numbers in parenthesis include the
probable but somewhat doubtful cases.

progeny of "WG9-C10, is decidedly earlier than the adjacent lots.
Lot 25 also appears early, however.

Tables 17 and 18 give a direct measure of earliness, relating to the
primary inflorescence alone. The clusters visibly aborted were in
general relatively far advanced, and those aborted at the earlier date
correspond to decidedly early flowering; consequently the flowering
and aborted clusters are classed together as early. Chart 3 gives the
percentages for singles.

Here the data for the doubles show fairly consistent differences in
the number aborted at the earlier date, while the October totals are

[in]

254

MISCELLANEOUS STUDIES

less regular. There are contrasts similar to those of table 15 up to
lot 26, which is late, while the check lots 27 and 28 are early. The
singles show the type differences very strikingly throughout lots 1
to 20, while lots 21, 22, and 26 give less positive indications of the
presence of the early factor.

Table 19 gives the numbers of singles flowering, in primary in-
florescence or elsewhere, by November 3, when growth had practically
stopped. The indications are in general the same as with the data
already discussed, with better evidence than usual that lots 21 and 22

TABLE 18
Same as table 17, for doubles.*

Row

Lot

Aborted by
July 29

Flowering or
aborted by
Oct. 10-16

Lot

Aborted by
July 29

Flowering or
aborted by
Oct. 10-16

1

1

15

30

2

6

22 (23)

2

3

23

33 (34)

4

21

35

3

5

12

29 (30)

6

16

30 (31)

4

7

17

30

8

8

22 (24)

5

9

25

41

10

24

41

6

11

25 (26)

40 (41)

12

23

40

7

13

16

28

14

22

31

8

15

27

37 (39)

16

11 (12)

29 (32)

9

17

20

35

18

20 (21)

32 (33)

10

19

21

41 (42)

20

21

42(44)

11

21

22

35

22

18

27 (28)

12

23

16 (17)

27 (28)

24

10

16(17)

13

25

17

25

26

9

15

14

27

21

29 (30)

28

20

25 (27)

* See note to table 17.

possessed the early factor. The mean time of flowering is irregular,
but shows some effect of the earliness factor. Lot 26 is late as to
number flowering, but early as to mean.

Table 20, for doubles flowering by August 1, no doubt gives more
reliable means; these means disagree with our scheme only in lot 26
and perhaps lot 22.

According to tables 17-20, the fewer-noded check parent of each
check row has usually given the earlier progeny. In fact, the agree-
ment of parental and progeny differences, throughout the cultures, is
decidedly remarkable. It is unfortunate that the later parents were
always placed in the east half of the row, especially in view of the
fact that there was indication of important differences in soil and

[112]

MUTATION IN MATTHIOLA

255

TABLE 19

1911, field; plants transplanted from greenhouse. Time from sowing to
emergence of earliest corolla. Singles.

Row

Parent-lot
internode
mean

Lot

Parental
internode
number

Progeny flowering
by Nov. 3

Lot

Parental
internode
number

Progeny flowering
by Nov. 3

Number

Days to
flowering

Number

Days to
flowering

1

29.60

1

29

7

147.14

2

32

7

128.57

2

21.40

3

16

34

91.94

4

20

33

105.45

3

5

25

26

119.46

• 6

27

25

104.08

4

49.57

7

46

30

103.13

8

54

24

105.67

5

9

21

30

91.73

10

21

34

91.12

6

27.33

11

22

33

98.85

12

25

27

108.30

7

13

34

18

100.67

14

41

13

120.62

8

28.50

15

27

17

112.94

16

29

18

118.00

9

17

33

30

100.27

18

35.

18

109.67

10

19

36

34

97.35

20

37

21

117.81

11

42.56a

21

42

25

129.36

22

45

25

121.76

12

23

49

11

122 . 00

24

51

14

151.57

13

25

55

15

136.40

26

13

121.08

14

47.80

27

46

10

159.40

28

56

8

162.50

a This parent-lot value does not applv to lot 26, which consists of progeny of
WG9-C10 itself.

TABLE 20
Same as table 19, -for doubles flowering by August 1.

Row

Lot

Progeny flowering by Aug. la

Lot

Progeny flowering by Aug. 1

Number

Days to flowering

Number

Days to flowering

1

1

38

90.26

2

33

91.03

2

3

36

80.22

4

36

80.00

3

5

39

84.46

6

39

84.10

4

7

36

81.28

8

31

84.32

5

9

42

75.86

10

41

76.59

6

11

42

77.90

12

40

80.25

7

13

37

84.32

14

35

84.80

8

15

46

83.87

16

33

85.21

9

17

37

80.43

18

34

84.12

10

19

42

78.24

20

39

83.85

11

21

39

85.95

22

30

87.93

12

23

33

89.03

24

26

88.85

13

25

32

88.56

26

21

89.05

14

27

35

88.97

28

29

90.28

a Only 48 more doubles altogether flowered by November 3, and 25 of these
were in the even-numbered lots 20 to 28.

256

MISCELLANEOUS STUDIES

probably in the incidence of disease, favoring the plants in the west
half. The internode data of 1910, however, show a similar tendency.
Small genetic differences are suggested, though it would be remarkable
if they were so uniformly present in these plants of a single line of a
usually selfed species, descendants of parents and a common grand-
parent grown under glass.

If such differences exist in the race, conceivably some combination
due to crossing might simulate an early mutation. The evidence as a
whole, however, does not favor such an origin for our early type ; it is
widely divergent from the Snowflake type, and seems to depend on
a single main factor difference from Snowflake.

TABLE 21
Cultures of 1912. Ancestry and parental data.

Parental data

Lot

Parent

Probable type

Days to

Inter-

Seeds sown

flowering <*

nodes'1

1

WS1-W216

Snowflake8

120.5

38

15

2

WG9-C10-W6

Early

116.5

33

15

3

WL10-W>2

Snowflake

139.5

51

15

4

WL10-W23

Snowflake"

120.5

38

15

5

WS1-W21

Snowflake

141.5

57

15

6

WL10-W214

Snowflake"

126.5

38

15

7

WL10-W27

Snowflake

145.5

54

15

8

WG9-C10-W8

Early

129.5

45

15

9

WS1-W212

Crenate-leaved3-1'

119.5

34

7c

* Suspected before testing of belonging to the early type; first parent also
tested in 1910.

b A heterozygote between the crenate-leaved and Snowflake types.

c Probably open pollinated.

d All the parents grew in the same house at the same time.

The essential feature of the supplementary cultures of 1912. since
no seed of WL10 remained, was a test of two pairs of early and late
progeny of "WL10 (lots 3 and 4, 6 and 7, table 21), in comparison with
two control lots — one (lot 2) from a known early parent, descended
from WG9-C10, and one (lot 5) from a late descendent of WS1.
Incidentally, WS1-W216 and WG9-C10-W8 were retested, and the
few available seeds of WS1-W,12 were used to test that phenotypically
early parent.

The results are given in tables 22 and 23 and chart 4. The very
low individual from WS1-W216 came from a very weak embryo, and
should be disregarded; the exceptionally high general range of this
lot, which was also visibly behind all others in development, was prob-

L114]

257

TABLE 22

Cultures of 1912. Number of main-stem internodes 'below first flower-bearing
node. Frequency distributions for singles.*

1 Gen. 1

WSl

WG9

WL10

WSl

WL10

WG9

WSl

Ancestry •< Gen. 2

W216

CIO

W22

W23

W21

W214

W27

CIO

W212

[ Gen. 3

W6

W8

Internodes
18

It

19

1

20

1

21

22

1

3

23

24

2

25

26

It

i

i

4
1

1

i

1

3
3

1

27

2
1

2
1

28 ..

3

2

1

U
It

29

1

2
1

30

31

32

33

34

35

2t

* See note a to table 5.

TABLE 23
Cultures of 191S. Same as table 22, for doubles."

[ Gen. 1

WSl

WG9

WL10

WSl

WL10

WG9

WSl

Ancestry -i Gen. 2

W216

CIO

W22

W23

W21

W2 14

W27

CIO

W212

1 Gen. 3

we

W8

Internodes
12

1A

13

14

15

16

17

1

18

1
1
1

2
2
1

19

20

1

21

1
1
1
1

2
3

2
2

i"

2

i"
i

22

1

23

i

5

1

It

3
3
2

2"
3

i"

3

1

24

25 ..

26

27

2

3t

28

it

29

30

1

31

1

a See note a to table 5.

[115]

258

MISCELLANEOUS STUDIES

ably due tu some cultural accident, perhaps to an excess of moisture
in this row of pots.

The lots of plants may seem rather absurdly small for their pur-
pose, but the uniformity of development here, with the marked normal
divergence in internodes of the types in question, seems to justify a
fair degree of confidence. Ten plants here were probably worth fifty
in the field.

31
30
2<)
28

27
20
2,5
21
23
22
21
20
19
IS
17
If.

c,

r

ingl
oubl

68
68

-

-

1

1

—

.__

_ — -

.__

i

1

• —

. —

1 — <

^

•~~"

~— ~

<

i

W,16 W6
WS1 CIO

W22 W23

i i

Wil

1

WS1

W,14 W,7

i

W8 Wil2
CIO WS1

wiio

wi,io

WG9

Ancestry

WG9

Chart 4. Cultures of 1912. Internodes: parental values and progeny means,
shown as in chart 1. The true parental values are twice those indicated by the
ordinate figures, which apply directly to the progeny values.

This test, with that of 1910, shows very positively that WS1-W216
was only phenotypically few-noded. Evidently WG9-C10-W8, the
parent of field lot 22, really carried the earliness factor, as was some-
what doubtfully inferred from the field results; the five progeny of
WS1-W212, on the other hand, though from a fewer-noded parent,
have values that make the presence of the earliness factor improbable.

On the main point at issue the evidence seems satisfactory. Neither
of the two very early and few-noded progeny of WLIO represented

[116]

MUTATION IN MATTHIOLA 259

shows in its progeny any evidence of belonging to the early type ; the
means are slightly lower than for the many-noded sibs of these parents,
but far less so than with the parents descended from WG9-C10.

We conclude, then, that WG9-C10 was probably a monohybrid,
and that the early-bearing gamete entering into its composition was of
unknown but presumably mutative origin.

Most of the extracted late or many-noded parents may now be
selected with practical certainty. "WG9-C10-C8 and Cl (lots 5 and 6
in the 1911F cultures) and WG9-C10-M7 and M8 (lots 13 and 14)
were genetically very similar to the check parents, as has already been
concluded for two of them from the greenhouse cultures ; presumably
they were pure Snowflake.

The data for WG9-C10 itself (lot 26) seem to indicate that the
results from the last eight lots are of very doubtful value ; still, they
show, especially in the original individual records, some evidence of the
earliness factor which must be present in part of the individuals. The
poor and slow germination of the old seed available may have had an
important influence on the result ; many of the early embryos may have
been non-viable, and the seedlings may have been weaker than those
from fresh seed. The 1911 data and observation of the plants in the
field suggest that WG9-C10-W7, W3, and W10 (lots 23, 24, and 25)
are the only remaining extracted late parents, WG9-C10-W5 and W8
(lots 21 and 22) carrying the earliness factor, as the four parents just
preceding them in the cultures obviously did. Tables 22 and 23 con-
firm this conclusion for WG9-C10-W8.

It is presumably impossible to make a positive separation of the
parents homozygous for the presence of the early factor. The green-
house data suggest that WG9-C10-M4 was a pure early individual;
the field data (see lot 9) agree, and suggest that WG9-C10-M9 (lot
10) and perhaps WG9-C10-M6 (lot 11) belong in the same class.
WG9-C10-C2, C5, and CIO (lots 3, 4, and 40) 12 were all evidently
heterozygous. Of the parents grown in house W, it would seem that
only WG9-C10-W11 (field lot 19) was homozygous early. We have,
provisionally, for the available single progeny of WG9-C10:

House 0 House M House W Total

Pure early 0314

Hybrid early 3159

Pure late 2237

20

12 Statistical data given for the last only for the 1910 cultures, not for this
field lot.

[1171

260 MISCELLANEOUS STUDIES

This corresponds well enough with the monohybrid expectation of
5 : 10 : 5 ; in fact, the deviation is just such as would be expected if there
was occasional cross pollination of the unprotected flowers of WG9-
C10 from Snowflake plants. The large proportion of evidently pure
late parents is strong evidence for the monohybrid nature of "WG9-C10.

The proportions of the two types among the doubles can only be
estimated. The 1908 data suggest that 5 of the 10 doubles there
reported were early ; this number, with the 13 singles so classed, makes
a total of 18 early-type plants out of 30. The ratio is slightly nearer
to 1 : 1 than to 3:1, and the former proportion would suggest the
peculiar type of inheritance found with the mutant types yet to be
described. The evidence of the 1910 distributions, however, shows that
the early type largely predominates in the next generation with both
singles and doubles, and apparently this is true even when we exclude
the progeny of the one parent classed as pure early.

The early factor can be positively detected only by progeny tests.
No test has shown the presence of this factor elsewhere than in WG9-
C10 and part of its descendants. WG9-C10 produced the early and
Snowflake types among 20 single progeny nearly in the typical mono-
hybrid proportions: Inspection of the double progeny in two genera-
tions suggests similar or possibly somewhat lower proportions there.
A vicinistic origin for WG9-C10 is improbable. Presumably, then,
the early type arose from Snowflake by a single factor mutation, the
dominant mutant factor being inherited without special complications.
We shall now consider certain apparently mutant types which are
characterized by peculiar genetic behavior.

2. THE SMOOTH-LEAVED TYPE

This type was first observed in the cultures of 1908 (table 1) and
has occurred frequently in later cultures (table 3). It is perhaps the
mutant type of most frequent occurrence among progeny of Snow-
flake or early parents; 2410 unselected progeny from house-sown seed
of such parents (see table 28) included 28 apparent mutants (14
singles, 11 doubles, and 3 undetermined), a mutation coefficient of
1.16 ± .15 per cent.

As grown in the greenhouse at Ithaca, this type (fig. 7, tables 12
and 13) was often many-noded, with correspondingly late flowering.
Its most striking peculiarity, shown especially by young seedlings and
not evident in the figures, was a lack of buckling between the veins

[118]

MUTATION IN MATTE 10 LA

261

ri <rr ••'

05 O O

PJ rf C

to o

III

cc o>

«"*" 03

* ^ »

g«i
' g &

,

on;

* s

°1S. °'

« o

O M,

5-2

00 CD

sJ!>tototototO2:>co;> w to to £ J> w to to

Kt-osososascsgc-gc-. as cs GS g £-• as 35 as
9 —o 5 o o o-§ —g — o 7 erg — « crp

1 ^icJocoJ,t IS co 4X1^

g. ^J^ P- Er oy
fD in

Parents

^ CO CO O CO CO — ^ CO CO CO CO CO !— CO CO CO CO CO

Q

»-> CO CO 00 h- CO h-> CO i-> (-1 i-» O *-> O I-*

K W ^ W^ K W ^ gc ^ o, ^
to" to"

H- ' h- '

*. ){X

1
1

tO Oi ts5 I-1 >— to >~*
O tOlOCDCntOO Oi h- H-" tO^ ^ v} ^
CO O W^cn O O O O5 00 CX) O^ JO vO Co

Oi tf» ^ ££> i~*
O C) >— • "-» Cr^
•^ -^

8?

•

L

Cr

to ootooooo o» co to oco oj o o

Undeter-
mined

H
o

?

n

I—1 tO I— ' h->
M to tO M £- i— CO h-> Oi Cn H- ts3 tO O>

CO O O 00 >— ' CO *»• CO Cn OS CO O» H- > CO Oi

Deter-
mined

<amined

: :::.:!—'

5'

H

o

o

0
ft

^

:::::*- 00 O5 to O O> 4^ to :

B

tr

i

(B

IB

<;

9
PL,

1

OS CO !-• h-> 00 h-
O OOOOOCO h- ' ?D t— ' H- 'tO 00 O5 OO

as "co 'to
^ i5 -S

>

2
P"

9

e

00 l—COtO 4*. CO h- IO
I-" -~J (-- CO rfi. Oi tO *«• tO O tO 00 bO O :

'00 "OO^'oo'tO ^ 'CO "h^ T3

-S "^^B^-S S c3 w M

SP
5'

N_

!T

<l H-i COh-ih-irfi. CO H-i H-*

^> O^IOOOOh-'^l «5 O> CO h^-00 CO Oi :

00 'H^'Pc;? ">£ 'H- 'to
CO Cn CO O tO 00 O

H

o

2

Snowfla

0

I— ' h- '
O5 h-> h-Kl CO H-> Oi M tO tf»- ^1 >-« tO 4>
I-* 00 00 H- M ^J h-i 4^ OO O Oi <l h-i tO »*>•

t-> H- « 1— ' -^j'4». H- ' 1— ' 00 CO rfk OOI-« 4^

O3 CO CO tO O CO OS O H- CO >-" CO OS

>

^- ^

CO h- ' H-> t-> O 4^- CO I—1 tO I— ' O I—
4^ r-> CO-<Jrf^ H-itO _!-*

CK"
S?

n <

J

-

QQ

[1191

262 MISCELLANEOUS STUDIES

of the leaves, and of general convexity of the upper surface of the
leaves. Mature plants developed under favorable conditions in the
greenhouse closely resembled Snowflake; the leaves, however, were
noticeably brittle, and the dry capsules so brittle that it was often
necessary, as it was not with Snowflake, to shell the seeds individually.
Probably the fibrovascular system is in some way defective ; Oenothera
rubrinervis, which is also brittle (deVries, 1906, lecture 18), has
thin-walled bast fibers.

In the field cultures, both at Ithaca (fig. 5) and at Riverside, under
conditions less favorable on the whole to the initiation of flowering,
this type (fig. 8) differed much more widely from Snowflake. Flower-
ing was excessively delayed, and the plants often remained low, with
few branches, and rosette-like, with thin, rather narrow leaves. Small
brown dead spots, possibly due to excessive transpiration, occurred so
frequently on the leaves as to constitute a good diagnostic character
for the type. Another peculiarity observed in the field is a reflexed
position of the tip of the young leaf when first visible — Snowflake
leaves being completely erect from the first.

In the 1914 cultures, with better development than in other field
cultures, some smooth-leaved plants (figs. 9 and 10) were again more
like Snowflake, though later and evidently more leafy.

Six smooth-leaved parents have been used in progeny tests, three
of these being apparent mutants and three being Fa progeny of two
of those mutants. The results are presented in tables 24 and 25 ; these
tables require a brief explanation, which will apply also to the similar
tables for other types.

For the plan of the new pedigree numbers here used, see ' ' Methods.
The initial plants of a series are designated as the Px generation in
the tables, their progeny as F,, etc. In table 24 the cultures are
arranged according to their generations and their pedigree numbers
under each generation; the smooth-leaved parents (Pt or of the P,
type) are given first, followed by the extracted Snowflake parents.
In table 25 "good germination" indicates that in all lots included
(taken as grown, not as summed by parents in table 24) the number
of plants determined exceeds 50 per cent of the number of seeds sown,
and vice versa; the weighted mean percentages obtained by dividing
the total numbers of plants by the respective total numbers of seeds
are given for each table in a footnote.

All six smooth-leaved parents (tables 24 and 25) gave mixed
progeny, part smooth-leaved and part Snowflake. The surprising

[120]

MUTATION IN MATTHIOLA

263

fact is that the parental (smooth-leaved) type appears not in three-
fourths of the progeny, but in only about one-fourth.

The extracted Snowflake parents tested behave like pure recessives,
showing no influence of their smooth-leaved ancestry. Only the
aberrant ratio seems inconsistent with the assumption that the smooth-
leaved individuals tested were ordinary heterozygous dominants.

The relatively weak growth of this type and the apparently poor
germination of the seed produced by it suggest that normal segregation
may be masked by selective elimination. Possibly the smooth-leaved

TABLE 25
Smooth-leaved type: heredity. Summary.

Progeny

Plants

Parents

Cultures

Seeds

Total examined

Smooth-leaved

Undeter-
mined

Deter-
mined

Number

Per cent

All smooth-

leaved

Ithaca

304, 217

7

156

40

25.6 ± 2.4

All smooth-

leaved

Riverside

196

1

78

23 (24)

30.8 * 3.4

All smooth-

leaved (6)

All

500, 817

8

234

63 (64)

27.4 * 2.0

All Pi smooth-

leaved (3)

All

244, S17

3

115

32

27.8 * 2.8

All Fi smooth-

leaved (3)

All

256

5

119

31 (32)

26.9 ± 2.8

All smooth-

Germination

leaved

good

293, 138

8

187"

55 (56)

29.9 ± 2.2

All smooth-

Germination

leaved

poor

207, 79

0

47a

8

17.0 ± 4.4

All Snowflake

(5, Pi and F2)

All

208, 50

2

173

0

0

* Respectively 63.8 and 22.7 per cent of the numbers of seeds planted.

factor is lethal when homoz'ygous, as is often the case (Muller, 1918)
with dominant mutant factors in Drosophila; the data for germi-
nation, however, indicate that two-thirds of the mature embryos can
hardly belong to the mutant type. "We might expect, in view of the
weak growth of smooth-leaved plants, that partial elimination of
heterozygotes would also occur. That this is the case is suggested,
though the numbers are small, by the lower proportion of the mutant
type with poor germination (table 25; see also tables 39 and 40) ; it
should be noted, however, that transferring the first lot of table 24,
the only lot between 50 and 73 per cent, to the "poor" total, makes
the percentages practically identical.13

See also table 2 and the second paragraph under "Occurrence of Mutants."

[121]

264 MISCELLANEOUS STUDIES

In connection with the question of lethal action we must consider
the inheritance of doubleness of flowers. Snowflake seed regularly
gives a mixture of singles and doubles, about 53 per cent being doubles.
The doubles, which are totally sterile, are probably (Frost, 1915) pure
recessives (dd) for a single-double factor pair. The singles are always
heterozygous (Dd) ; crosses with pure single races (Saunders, 1911)
show that the approximately 1 : 1 ratio and the failure to produce pure
singles, with self pollination, are due to the fact that all the functional
pollen is doubleness-carrying (d). The excess of doubles over 50 per
cent has been explained by Miss Saunders (1911) as due to hetero-
zygosis of the singles for two linked complementary factors necessary
to singleness, and by the present writer (Frost, 1915) as due to lower
viability of the "single" gametes or embryos. The absence of func-
tional single-carrying pollen is apparently due to a lethal factor acting
after separation of the microspore tetrads, since the tetrads themselves
appear normal.

In any consideration of factors linked with the single-double pair,
this semisterility of the pollen must be remembered. For example,
any dominant factor completely coupled with D in pollen formation
would be totally absent from the functional pollen, and the zygotes
produced by selfing would show directly the strength of linkage in the
ovules.

The available data for the smooth-leaved type (table 24) are far
from constituting an adequate test of linkage, but they suggest that
the factors are independent. Certainly no high degree of linkage is
indicated by the totals, nor do the detailed data suggest that smooth-
leavedness is coupled with singleness in some parents and with double-
ness in others.

We must admit that the peculiar inheritance of this type is not
yet positively explained. Evidently larger cultures are needed, and
crossing with the Snowflake type and with other commercial varieties ;
cytological study may also be required. Certain comparisons and
speculative possibilities deserve mention, however, especially since the
types yet to be discussed furnish additional evidence bearing on them.
We may compare the smooth-leaved and double types, as follows :

DOUBLE SMOOTH-LEAVED

1. A rare mutation of pure single 1. Apparently a common mutation

("normal")- of pure Snowflake ("normal").

2. Recessive; extracted recessives 2. Apparently dominant; extracted

are sterile mutant-type plants. recessives are fertile normal

plants.

[122]

MUTATION IN MATTHIOLA

265

DOUBLE

3. Homozygous dominants not pro-

duced by hybrids, because func-
tional pollen carries recessive
factor only.

4. Kecessive (mutant) type the
more vigorous.

5. Dominant factor or another fac-

tor very closely linked with it
is incompatible with formation
of functional pollen.

6. Eecessive type exceeds the ex-

pected equality by about 3 per
cent among some 7000 indi-
viduals.

SMOOTH-LEAVED

3. Homozygous dominants perhaps
not produced by hybrids.14

4. Eecessive (normal) type the
more vigorous ; difference much
greater than with single and
double.

5. Eelation of dominant factor to

viability of pollen not yet de-
termined.

6. Eecessive type exceeds equality

by about 23 per cent among
234 individuals.

The most probable hypothesis for smooth-leavedness, then, would so
far seem to be essentially the same as for doubleness — complete elimina-
tion of the weaker type in pollen formation, and partial elimination in
embryo-sac formation. Reciprocal crosses with Snowflake are obviously
necessary; as we shall soon see, three of the other mutant types have
already proved to be carried by both eggs and sperms.

The case of Oenothera lata (Gates, 1915) suggests the possibility
that the smooth-leaved form might arise by reduplication of a chromo-
some. With ordinary 0. lata the pollen is sterile, but pollination by
0. lamarckiana gives about 15-20 per cent of lata. This deficiency of
lata individuals is due, it seems, to a frequent loss of the extra
chromosome at meiosis in lata ovules, with a resulting formation of
more than 50 per cent of seven-chromosome (lamarckiana) eggs.

If the smooth-leaved type originates through duplication of a
chromosome, we might suppose that other types of similar heredity
involve other pairs of chromosomes. The apparent parallel with
0. lata, which Bartlett (1917) has noted, was long ago suggested by
the data, but with at least two or three types to be described linkage
phenomena have seemed to conflict with this interpretation. Possibly
different processes have produced different mutant types as with
Oenothera; as we have considered types suggestive of 0. rubrinervis
(early) and of 0. lata (smooth-leaved), we may consider next a form
which in appearance is remarkably suggestive of 0. gigas.

14 This possibility is only suggested by these cultures, but it becomes highly
probable when the data for other types are considered.

U23]

266

MISCELLANEOUS STUDIES

t

(

38

J

fl

(N OOOOO OOOO^H -HO-H>-iO TfOOOOO O O

P to

a

S3-Sd-2- ^ -H 3" ^ *~I.§S d-§- rt

"*•

«C O CO ••# 00 i-H «C Tj< CO 00 OC »C t^ O t> CO -Ht^Ttt-^OJ-^ O »C

C^J i-H -HO CN »C 1-1

i-l •-!

c.
M

0)

^ _ _ _

1

i

3

Q

OS S S SS §S £

:

Q

<NOiC<Mt^COCOT*<COcOcC'«J<CO»t(N-H O3iM-Hi-i»CO5 O 00

3

m

"5

.

He <N S?f^2S

1

00

•r.

CC ^5 CO C^ ^H iO C^ — t^ ^1 C^ ^^ "*^* tO iO C^ C^ *O CO CO ^* ^O ^^ Is*

-H Tf 1-1 I-

1

(••v ^^^^ v^x—S. x s^D X— s^D ^H /-— «s ^— V ^D -^~N CO T1*

^,

"^

CX5C^1iOt*C^C^OCGiCOCOC^IOiOt^CCC^ OOOCOC^IO^O CO O

5

CO ^H f-H G5 M W5

i-H

•£

w

=

~3

—

o

v

I

•

^, 2* ^ S

=

s

i

\

3

^COTPCOCOt"** — '~ — i~. — ^HQO

'Z

V.
3

A

£

1

O

Q

"^ c^ co *-O ^^ ^H cc Is* co co ^^ ^^ ^o ^o ^^ ^5 t*» co ro c*^ co ^^ c^ ^

r-t Tf i— I I>-

13

=,

J

~.

9

a

e

'-™^ x— v

P

fl>

x — xx-*«, x — ^ x-*\^"*s.-- — s-— -^ ^^ Is1*

r

0

li

=

SSG- G SSS-SG- 2!- t

_^

^

£

COOC<>i-Hr^-H-HC<|iMCO-HiC»Ci-l<NC^ ICCOCOOCOIM O O

f

(N CO 1-1 t-

^

1

1

11

co co t^ t^ c co r^ co as Tti oc t>- oc o -H t>- ooo-Hcoci'f t^ »c

be

«J

'§

11

(N CO

r:

^>

m

•s.

£_

c"

$•0

^

8

o

1.1

<N^O-HOOOOOOO<NOOOO ^ooooo -o o

•

|

H

c g

ts

* " * ~

V.

Ci -H ^

j/:

* *-. »-H CO CO

i

C*l "^-OQ Oo *^?* ^^t* ^<t" ^^t" ^^5" ~ -v^t"'^?*^:t"^t" ^t^ OC ^J ^"•J'^f'OQ ^^C^ CO ^4 ^?*
^«l ®") "^t" ^^^i O> 00 ^J O> ^4 Tf^ ^J ^} ^J C^ C*~J ^^ ^^* ^? ^3 *^» Q$ Ci C^ ^^ C^
i-l " W (N rf EO CO -^

Tt^CO
i-( 1-4

OS I CO CO CO CO CO CO CO CO CO CO CO CO CO O CO CO CO CO

71

|

"""' ^''lAJi'i^'JiJi11 ''''i '

~.

O

OS^OJOSOJOSOCJOSO: OSOSCSSSCSOSC:^: CiOCSClCS — 0

O

39

V

OOC 00 J.

COCOiMC^^g3,- ^

o

a

%

i— IOOOSOJCO''t<^fiC»C'CCOJOCOt>-f>-Cs4 OOOOOSCsfeS"*^ t^1^

llllllllliT Till jiiii^g- i cc

<0

T" O5c3c3c3c3c3o3c3^^3^3rf^c3c3c? rtc5rfc3o3 \ > — c3 ^

00

n
CM

* c5SSc^c^SSc5c^c5c5c5c^c5c5c5 cSc3cS?i?5'<J5 *§

-

oB

[124]

MUTATION IN MATTHIOLA

267

It should, however, first bf noted that, as will appear later, phe-
nomena of apparent linkage in the case of certain other types (crenate,
slender, and narrow) suggest that these forms commonly arise from
Snowflake by segregation rather than by immediate mutation. The
obvious objection to this hypothesis is the fact that the apparently
mutant types seem to be dominant to the ' ' normal ' ' or Snowflake type.
This objection can be met by assuming the presence of dominant in-
hibiting factors in the Snowflake parents that give apparent mutants.15

If the apparent mutants of the smooth-leaved type are thus pro-
duced by crossing over in a set of balanced factors, the lethal ' ' balanc-
ing" the smooth-leaved factor itself may be distinct from that which
sterilizes the singleness-carrying pollen. In considering the results
here reported, therefore, we must always bear in mind the possible
presence of several unidentified lethal factors. If the apparent absence
of linkage between the smooth and double factors is not misleading, we
must suppose that these factors are carried by different pairs of
chromosomes; considerations advanced by Muller (1918, pp. 479^82),
however, make it rather probable that the commoner types of apparent
mutants here discussed are all due to factors carried by one pair of
chromosomes, the pair containing the factor for doubleness and its
normal allelomorph.

3. THE LARGE-LEAVED TYPE

A double of this type probably occurred in the 1907 cultures,
though its appearance attracted so little attention that no record was
made. In the field cultures of 1911 (table 3) several individuals sug-
gested a gigas type, though there seemed to be intergradation with
Snowflake. In the 1912 cultures a single with leaves "long, rather
narrow, thick" developed normally and produced an abundance of
good seed; from this individual (28a) all cultures of this type are
descended.

This type is stout and coarse throughout, and late to flower. The
leaves are strikingly long, thick, and rigid, though as a rule relatively

" A letter suggesting this explanation was received from Dr. Muller soon
after the same idea had been outlined in the "General Discussion" section below.
Dr. Muller kindly gave further attention to difficulties at first encountered by
the present writer, materially assisting in the formulation of an apparently
tenable form of the hypothesis. Since, however, this scheme may seem ' ' far-
fetched ' ' and unduly complex, it appears desirable to leave the original discus-
sion of the individual types substantially unchanged. When the difficulties
encountered by the assumption of frequent true mutation have been more fully
presented, the need for some such addition to the scheme will be more evident."

[125]

i>r>s

MISCELLANEOUS STUDIES

narrow ; under unfavorable weather conditions the flowers are often few
and defective, while the leaves are resistant and long-lived (fig. 11).
Figures 12 and 13 show well the coarse leaves and lateness of well
developed large-leaved plants in the 1915-16 cult'ures, the plants in
the latter figure being several weeks the older.

The results of the progeny tests are given in tables 26 and 27. All
the twenty large-leaved individuals tested have given mixed progeny;
the proportion of the mutant type, though much larger than with

TABLE 27
Large-leaved type: heredity. Summary.

Progeny

Plants

Parents

Cultures

Seeds"

Total examined

Large-leaved

Undeter-
mined

Deter-
mined

Number

Per cent

28a

1913, 1914,

& 1915-16

122

2

73

38(40)

54.8 * 3.9

28a-F, (3)

1914

120

2

40

14 (19)

47.5 * 5.3

28a-F, (12)

1915-16

288

2

190

76 (90)

47.4 ± 2.4

28a-F2 (4)

1915-16

90

0

54

25 (26)

48.1 =*= 4.6

28a-F, & F2 (19)

All

498

4

284

115 (135)

47.5 * 2.0

All large-leaved

(20)

All

620

6

357

153 (175)

49.0 * 1.8

Large-leaved

Germination

good

360

3

260b

115(131)

50.4 * 2.1

Large-leaved

Germination

poor

260

3

97b

38(44)

45.4 ± 3.4

Snowflake (1, F,)

1915-16

24

0

15

0

0

• Mainly from unguarded flowers; see table 26.

b Eespectively 72.2 and 37.3 per cent of the numbers of seeds planted.

smooth-leaved, approximates to 50 per cent, not 75 per cent, with little
indication of selective elimination with poor germination.10

Here plainly, as with smooth-leaved, no pure mutant-type parent
has yet been tested. Since this is also true of the other types, aside
from early, that have been somewhat extensively tested, and fifty-three
mutant-type parents in all have given Snowflake progeny, it is prob-
able that homozygous individuals of these types seldom or never
develop. The actual adult ratio with large-leaved is plainly not 2 : 1.
but rather 1 : 1, a fact that would suggest absence of the mutant-type
factor or factors from the pollen. The small trial cultures started
in 1917, however, show that the type is carried by both sperm and
eggs.

is Since hybrids are of the mutant type in appearance, the possible cross
pollination by Snowflake parents could hardly give Snowflake progeny with any
pure large-leaved parent. It may, however, have reduced slightly the proportion
of large-leaved progeny from heterozygous parents of this type.

[1261

MUTATION IN MATTE 10 LA

269

If we are dealing here with a type cytologically like Oenothera
gigas, or rather the triploid semigigas, abnormal distributions of
chromosomes may occur at meiosis, giving unpredictable genetic
results. There has been special difficulty, as the numbers of doubtful
individuals in table 26 suggest, in separating large-leaved from Snow-
flake, though in part of the cases the difference is extreme. Possibly
some of the doubtful individuals are genetic intermediates due to
irregular meiosis in triploid nuclei; such irregularities in division
(Gates, 1915) occur with Oenothera. Both cytological examination
and crosses with Snowflake are plainly required.

TABLE 28

Crenate-leaved type: numbers of apparent mutants and association of the
type with singleness of flowers.

Culture

Progeny of Snowflake and early parents

Total
examined8

Crenate-leaved

Single

Double

All

Coefficient of
mutation

1908
1910
191 IF, seed house-
sown
All above
All unselected

725b
338

2072
3135
2410

6

3

13
22
16

1

0

3
4
3

7
3

16
26
19

.97 =»= .22
.89 ± .32

.77 * .13
.83 ± .11
.79 ± .12

a See note b to table 2.
b See note c to table 1.

4. THE CRENATE-LEAVED TYPE

This type (tables 1 and 3) is one of the three aberrant types of
most frequent occurrence in the cultures here described, having con-
stituted (table 28) about .79 per cent of the progeny of Snowflake
and early parents. A large majority of the individuals have been
singles, as table 28 shows. If the apparent mutants are produced by
some process of segregation of factors, evidently the crenate and single
factors were usually coupled in this material; if they are produced
by immediate factor mutation, or are individually due to some change
in a particular locus, evidently that locus is linked with the single-
double locus and the change is more frequent in the single-carrying
chromosomes; and finally, if they are due to reduplication or loss of
a chromosome, the apparent linkage remains to be explained.

The margins of Snowflake leaves vary from entire or slightly
sinuate to coarsely and irregularly dentate or serrate, this character-
istic being subject to much environmental modification and varying

[1271

270

MISCELLANEOUS STUDIES

i

<

" 00

> £

C<1 CO CoS" CD" C1!

j

: -

S»^ -^.X S^.XS^^ N^^ V ^

j

>«

«— ' O O »— i Ol >— iC^I^HC^ CflO ^f CO >C

=

G.S2- £3 d- £i ** i2. &

<

O5OC- »-i O 00 Cli-HC^CSO-HO >C C^ OC

— O4 04 CO ~. — CO ~~ ~~ ~~ CO • ~ ""•

(N

0

^^

M

O

•*-"•«. >^^», *i^*v /^*s C^

|
1

3
g

§ 5- S S d

c
c

Q

1— (

.

^s

"g1

•C t>- C^I CO

bo

o -^ o d co i— i *c co c^ c co ic oo oo »c

CO

jj

--v 00

|

a

^~, ^-v^~. Id <N

OS CO 1C CO >— I

*

**<

** O5 •* O5 CO 00 CO »C 1C 1C ^ O5 CO OO CO

(N CO •«* IM

a
o

M

I

—

s-~* Tf 1C

O

IB

2

0

3

3

m
B

B

O

Q

O M O »-< CO c^FHC<I-H'-ii-<-* CO 00 CO

»-H CO

O

o

"5

C3 t^* CO ^ CO CO ^ ^O ^ Tt^ CO ^C O Oi O^

c

»-i d CO 00

bo

1

fe-S

1C Ol ^C C^ '"I O Ci OO t^- CO 00 Ci O5 00 ^

1

•g.s

OS

i— 1 i— 1 i— t CO

5

o

11

O C4 NO"* O O O ^H O O O "* CO 00

f-

5s

* " *»( »"*

OC *H OO

m

T

(N >-H <N

•c~iN. os o" »c cToo -^oooooo oo oo o

COC>Tf CO I-H N*~i t>-iCNC^C<IC^C^ CD CD CD
i— 1 1^1 i— 1 i— I Tf ©i i— 1 i— 1 1C

£

^ •

°O ^O5

E

fa °^fa°^EiH °^ ffi K ffi ffi ffi ffi"*"1 ffi"""1

£

»-l N CO O "-I O »-l CO^tif^l^|^<^H--lrH-H CO~i— I CO"

0

Oi 05 ffi 05 0 O5 O5 5 O5OSO5O5C5O5O05O50>O5CS ^

03

Q> G)

|

- c ^^ 2ooos2 S - c

-— •" 1 1 t | 1 r^H •*

C«i

— /*• ^ ^Q _ u c3 — • • • ^3 13 'rt Tj

C^ C^ C^ ^^ ^ IN C^l !N (N C^ C^ C1 <M ^J

[128)

MUTATION IN MATTHIOLA

271

I

. ]

a

58

cJ '~S S" o^

,1

; a

S — •* *"— x N — x O^ ^^ ^H

C

l«

i-HO(MOOOO CO ^-'O^-' s-^
O

(M

&

if* C*} ^-^ t^ I-H /— x CM

3

C^ CO CM 00^ 5!,^, ?3

7

CO CO (N i-l i-Ht> CD CO OJ I-H 00 O3

CM i-l CO OO OS i-l i-l

co

•

^

jrf

eS

jj

^* ao ,~, e<? o^^ CD"

tc

.G

CM CM CM t^- CO CO i-H

£

3

O

' *• ' S S ' >* XV S -^^

0

a

C5 o t» i— i T-I CD 10 o; ic i— i co •*

GC

oi

o

s-^

IS

iO »^ CO

a

S ^N. ' V»^

oo

T^CC^COO»— I^H ^ ^O *O ^C

rH 1C

-2

0

I

£

^<

- — -

O"*r-iOOOO «3 -^OO O

i-H <N CO 00

E

T)

•

|

O>

rH • •

o

,— 1

3

£

I

•
C

g

Q

: : : : CO ^ ' ' ' '
iC(NcO:::: 1-1 01:: :

Tt*

g

c

"3>

»C (N rt< : : : : I-H "5 : - : :

c

T-I : : : : CM CM : : :

02

: : : : i— i : : :

1

C

i

II

t>» t"* 00 i-H CM t** CO 00 CO ^f OS CO

CO rH »C (N CO rH CM

i

M

* •

jg-C

3

O '

II

i-HOCOOOOO •* CO O O O

H

5s

i-H

00

•8

..^f

•
i
a

•* <N Q O 00 00 CO GOr-H*"
»CiMOi-<<N (NIMOOOOO 00

I-H CM i-H CM •* CM CO

CO CO CO

|

l-H I-H I— 1 |_|

1C >0»C S

3

i— I i-H i— Ir— ( r— 1 i— ( i— < _ _, i— IT— 1 _-

0

O5OO5O5O5O5O5 ^J33JJ O5O5 ^Jj

OOi^CNCOCOOS 5"cS i^*

H

0

|CM*O«5iC>C'C rtS OS OC*1!

i
PH

0 -O "O T3 "O "O T3 ^^ O.-3 ^bt3j-.^

c «

272 MISCELLANEOUS STUDIES

with the position of the leaves on the plant. In the crenate-leaved
type this character is much accentuated, as can be seen by comparing
figure 14 with figures 1 and 3 ; a warm greenhouse (fig. 14, upper line)
gave very marked serration, while a cool greenhouse (lower plant, and
also fig. 15) produced leaves much more nearly entire.

Under the much more extreme conditions of insolation, temperature,
and humidity at Riverside, this type was often much dwarfed in com-
parison with Snowflake (figs. 16 and 17; see also fig. 23). In general,
growth is weaker than with Snowflake and the stems more slender.
Buds and flowers are often produced in great abundance, but the
capsules are relatively few, small, and few-seeded. See tables 12 and
13 for internode data.

The progeny tests (table 29) show a slightly higher proportion of
mutant-type progeny than occurred with smooth-leaved. A striking
new feature appears for the first time in these results, the regular
presence of linkage, or an association simulating linkage, with the
single-double allelomorphs. Further, in all the four apparent mutants
tested the crenate factor seems to be coupled with singleness, while
among the sixteen F1 and F2 crenate parents there seem to be no
crossovers.17 We seem to be justified, for reasons just given, in
summing the progeny as in the tables. Two things appear at once in
table 29 : ( 1 ) there is a great excess of total doubles over the usual
53 per cent; (2) there is a much greater excess of doubles with Snow-
flake than of singles with crenate; (3) the supposed double-recessive
class (Snowflake double) is about two and one-half times as large as
the double-dominant class (crenate single).

Table 30 adds two features of special interest. First, there is good
evidence of selective elimination with poor germination ; compare the
remaining percentages with those for "Ithaca, field," "1915," "P1?"
and "Germination poor," and see tables 39 and 40; the only excep-
tional case is the low percentage for the thirty plants of 1915-16. It
would be surprising if the slow and weak growth of the crenate plants
did not lead to such a result. Second, there is evidence that the
crenate individuals are smaller than Snowflake even before germina-
tion. The seeds of crenate parents are less uniform in size than those
of Snowflake parents; small seeds are numerous, and even the larger
ones probably weigh decidedly less than normal Snowflake seeds.
With five crenate parents included in the cultures of 1913, random

17 With four of the parents the tests are obviously entirely inadequate; one
other, 22d-9, gives no indication of linkage among nineteen progeny.

[1301

MUTATION IN MATTHIOLA

273

• CD O O
JO J£) p3

•CJ TO
O O
GO
<-» PL,

tO H-

CD <I
(ft- - O5

h-i Ol Ol CO H->

Oo Qo

Ol h-i CO h-» Oi *>•

(-"vlCOtOtOOOtO h-O5

H-'WtOWl-'tOCOh-'l-'OOtOH-'OO

H-H-H-H-H-H-H-H-H-H-H-H-H-

[131]

274

MISCELLANEOUS STUDIES

samples of seed were sorted, and the smaller and larger seeds planted
separately.

Table 31 gives the data from this test. Here is practically con-
clusive evidence (see tables 39 and 40) that the smaller seeds much
more often contain embryos of the crenate type.18 Since the embryo
of a Matthiola seed occupies practically all the space within the seed
coats, it is evident that even as embryos Snowflake plants exceed

TABLE 31

Cultures of 1913. Crenate-leaved type: proportions from smaller and larger
seeds of crenate parents.

Seeds

Progeny

p t

Total

/"• 1 A

Size

Number

deter-
mined

vy rena te- lea veci

Snowflake

Other
types

Number

Per cent

22a-l

Smaller

21

6

4(5)

83.3 ± 12.9

0

1

22a-l

Larger

29

23

3

13.0 * 6.6

19 (20)

0

22a-5

Smaller

17

11

4

36.4 ± 9.5

6

1

22a-5

Larger

33

28

2

7.1 =»= 6.0

25

1

22b

Smaller

13

8

5

62.5 * 11.2

3

0

22b

Larger

36

31

4

12.9 =*= 5.7

25 (27)

0

22d-12

Smaller

30

24

17

70.8 =*= 6.5

5

2

22d-12

Larger

70

57

11(12)

21.1 ± 4.2

42 (44)

1

22d-15

Smaller

32

24

17

70.8 =»= 6.5

3

2(4)

22d-15

Larger

68

54

18

33.3 ± 4.3

34 (35)

1

All

Smaller

113

73"

47 (48)

65.8 ± 3.7

17

6(8)

All

Larger

236

193a

38 (39)

20.2 ± 2.3

145 (151)

3

All

All

349

266

85 (87)

32.7 ± 1.9

162 (168)

9(11)

• Respectively 64.6 and 81.8 per cent of the numbers of seeds planted.

crenate plants in size. This fact, obviously, is further evidence in
favor of the hypothesis of partial selective elimination of crenate
heterozygotes during embryonic development.

It may be worth noting that the 73 plants from the smaller seeds
include 6 (8) apparent mutants of other types (mutation coefficient
11.0 per cent), while the 193 plants from the larger seeds include
only 3 apparent mutants (1.6 per cent).

Before we can profitably discuss these data further, we must con-
sider the results from cross pollination (tables 32 and 33). The
numbers, though small, make it very probable that both eggs and sperms
carry the crenate factor. Further, it appears from series 20 that only
a small portion of the sperms carry this factor, as we should expect
from its apparent linkage with singleness. If homozygotes are non-
viable, the combined crenate percentages of reciprocal crosses should

18 The poorer germination of the smaller seeds suggests that the disparity
between the two lots of seeds in the proportion of crenate embryos was even
greater than the cultures indicate.

[132]

MUTATION IN MATTHIOLA

275

03 02

a »
o »

o;

p, Jf

sr*o co ;> to to to j> tr1 to to to to to K»TJ to to to

rv^ O tO •"" {i- CT" P tO *— * as £TTO CR !3J Q- ?? ' G* 3 f ;T P

"i 2 i >— ' o PJ tj^ P o o '-'• p.fM "*> CD o 5.4. CT^ cr P

V

!fl 1 if *f

I

r

Cn "* n a>- 5^° ft)
- — So Se

01

to o . - co o co E— • co co co co co co co co co co co

0

— . - Cn rf* co Cn Cn Cn Cn Cn cn rf* cocococo

-^

So

I

CJi OS Cv C^ * 3$ CT-

1

M

M <> CO CO O tO tO H-i h- Cn H- ' O5 tO <l *^

J-1 Cn O rf* tO O3 OOCn^JOOOCO CO *• O5 CO

I

X

§

m

CO i—" O O i— ' Cn OOOOOOO CnOi— 't^

S'l

H
o

2 S-

D-o

E.

*?

(D

H* tO I— '

30

i.

co to 'r'1—' on-'t-'i— >rf».i-»i— ' co osto

S'«-

5'

Cn OSh-vjOO COCC^tOCn^H^ ^COrf

S-7

n

a.

to : : tOOO COt-'i-'Oi-'O'. Cn: rfi.i-'

00

•S i i i i

a'

O O : : OCO CO O O H* O tO : : O

O

Drenat

o

"2 ' ' — " ' ' — "

§

7.

S

*--^ ^—s '. *^*s

cr

Q

a

5"

»

"<

1

1— 1
IO OOtOt-4 OS i-« h-> h- h- ' ts3 O CnOtt^t-1

, ^ ^.^^ ^^^^^^^ ^-^ , — ^

^»

3

i-« co co t-j os Cn

sr

»—

5

Cn •<> rf*. *-• co to
co co H-otoj* H- M en en ~j ^1 co Oi-«to<i

go

X * X-V y^^s .^-N .-^\ ^^^ -^X

5'

O5 -J hfe. h- ' 4^. CO OO
CO *J tO 00 — H-i ^-^

C5S i-« t-1 O *» I— ' . IO rfi.COt-'

O >yo en H-> to co oo o vi en to >*>• os co co to rf*

O

DG

5

"ps 1^ 'H^'H^ "en ^ ^ ^ 'Cn "So

O

|

F

n

x-

*-*

-3

h- ' l-»

H- tO H-> M 00 1— h-i h- CO H- VI Cn tO

p— ' to I— '1— ' CO H- ' P— ' rf^ h^ ' 00 CntO
tO LO CnCO4^CO)-'tO>^>tOa>tO

—

CO O OOOO5 tOOOOH- h-O •4XOCOH-'

**" C

)

^~^ y^^ x-^\ /*^^ y*^\

*o t

00 CO ^O V\ td

M S

[133]

276

MISCELLANEOUS STUDIES

exceed the percentage from selfed parents; the expected high pro-
portion with series 21, however, might well be realized with adequate
numbers and good germination.

In spite of the small totals, it is very probable that linkage similar
to that of the selfed cultures prevails with series 21. Where the
crenate type is the pollen parent (series 20) linkage ratios are on our
hypothesis impossible, since the eggs are all Snowflake and the sperms
all double; the data, however, though statistically inconclusive, sug-
gest that the excess of singles with crenate and of doubles with Snow-
flake is greatly reduced but not abolished.

TABLE 33
hybridization of the Snow flake and crenate-leaved types. Summary.

Progeny

Plants

Parents

Cultures

Seeds

Total examined

Crenate

Undeter-
mined

Deter-
mined

Number

Per cent

20aa, bb, & cb

1913

123

5

93

5(6)

6.5 * 1.6

20dc, ed, & ic

1914

163

0

14

0

0

20de,ff,gf,gg,&hd

1915-16

120

0

103

6

5.8 ± 1.5

All of series 20

All

406

5

210

11 (12)

5.7 ± 1.1

21aa, bb, & dd

All

75

1

25

2(3)

12.0 * 4.4

Snowflake par-

ents of hybrids

(5)

All

271,50

3

134

(1)

.7 * .5

If we may ignore the doubtful correlation just mentioned a fairly
adequate complete hypothesis for the selfing ratio is possible. Assume
(1) a gametic ratio19 of 5DC :ldC :1Dc :5dc, or 16% per cent of
crossing over; (2) non-viability of homozygous crenate (CC) ; (3) low
viability of simplex crenate (G'c), eliminating an average of 60 per
cent of this type ; and (4) coupling of D and C in all parents tested.
Evidence has already been presented for assumptions (1), (2), and
(3), except as to the intensity of linkage, while (4), as will be seen,
is not at all improbable.

Random fertilization under these conditions, excepting (3), would
give 26DdCc (crenate single) -(- lOddCc (crenate double) -f- 5Ddcc
(Snowflake single) -|- 25ddcc (Snowflake double). The other two
classes, 5DdCC and IddCC, would be non-viable pure crenate. Adding
assumption (3) gives the following comparison:

!» Representing the singleness and doubleness factors by D and d, and the
crenate factor and its "normal" allelomorph by C and c.

[134]

MUTATION IN MATTHIOLA 277

DdCc ddCc Ddcc ddcc

Theoretical ratio (n = 44.4) ...... 10.4 4 5 25

Calculated for n = 540 .............. 126 49 61 304

Observed (n = 540)20 .................... 125 51 57 307

This fit surely cannot be criticised, whatever may be thought of
the devices employed to obtain it ! With cross pollination the agree-
ment is fairly good in the case of series 20, which gives the only fairly
reliable data. We are assuming 16% per cent of crossover dC sperms ;
elimination of .60 of 16% per cent, or 10 per cent of the total, gives
.06%/.90 = 7.4 per cent expected crenate, as against 5.9 per cent
observed. Series 21 is supposed to have 50 per cent of C eggs in
the ratio 5DC :ldC ; elimination of .60 of this proportion, or 30 per
cent of the total, would leave .20/.70 = 28.6 per cent, against 12.0
per cent in the very inadequate material observed. An adequate test
of the hypothesis obviously requires large hybrid cultures, from
vigorous seed sown under favorable conditions for germination.

A scarcity of crossover crenate singles follows from the hypothesis ;
they constitute only one twenty-sixth of the total number of viable
crenate single progeny of crenate parents. No direct evidence indi-
cating that the crenate and double factors are ever coupled in singles
has yet been discovered.

If the supposed crenate mutants are due to immediate factor muta-
tion, however, it seems strange that the same locus is changed more
readily in a singleness chromosome than in one carrying the doubleness
factor, in a ratio similar to the linkage ratio of later generations.
If the apparent mutants are really segregates from a balanced-lethal
combination, the observed original coupling of crenate with single
might be an accident of sampling involved in the original choice of
material; other initial parents might give the reverse coupling.

5. THE SLENDER TYPE

This type is comparatively rare as an apparent mutant from Snow-
flake or early; the 3135 plants reported in table 28 gave only 4 (6)
mutants (2 singles and 4 doubles, 2 of the latter perhaps Snowflake),
a mutation coefficient not over .19 per cent. This type seems to occur
more frequently among progeny of crenate, a type similar in some

20 Omitting 29 plants classed as neither crenate nor Snowflake, which as
probably non-crenate should perhaps be added to Snowflake, and also 64 plants
(13 crenate and 51 Snowflake) with flower data incomplete. Complete data for
the total of 633 plants would plainly give a somewhat poorer fit, but this could
be improved by assuming a slightly greater elimination of Ccii zygotes.

[135]

278

MISCELLANEOUS STUDIES

1

I

j S

5? cJ "? S" £7 CO"

4

: a

*^^f ^ ^^* V s N^^ X_^

c

>«

OOOi-H T-H O IN O O <N t-H t-H

CO CO T-H t>» ic ^*

=

CO CO ^ CO i-H i-H

"•

O OS OS »Cl~- OOT-I t-H Tj< 1C OS <M l^

T-H t-H

0

v

s~* s-~* s-~* . — CO t-H

3

c

?

* C*i C^ t** CO 1C ^*

. co co co co 1—4 t— i

•£

p

'. ^^* *^^ ^^^ >*— X \»_^ **_^

o
CQ

Q

^ OO OO C^ Is* 00 t-H 00 ^* ^t* ^* C^ O5

1C 1C CO T-H 1C IM -* t-H CO

t-l 1-H

•

CO CO ^C 'C

c

N*^ N ' ^** ^^

in

• — — C< O OO C<> O T-H rt< O •***

1

— , s-^ ^. ^~, ^.^ s-^ ^

1

3

~* G- £i £i °° '"'t ^ JS

t^ t^ •* TH COO i-HCOOOOS OS O

T-H T-H T-H t-H CO T-H CO T-H 1C

>.

Ui
•O

3

s~^ s-^ ic : c5~ «N co co"

H

3

1

O

X

0

Q

O OO OO Tt« T-H t-l : CO T-H C^ h- C<1 1C

^ —, s-.s-^

•a

T^T Ct T-H t-H rfl CO

c

^ ' ^ V • . ^rf"S» X ^.^

02

OS OS O O "^t* ' ^f C*3 ^C O CO ^

T-H i-H i-H Tf CO

1

~ £

i— 1 CO ^f COO r— 1 rH OOb»TfHCO 00 1C
OO OO »CiM CO O •* •* CO O

I

11

T-H O1 IM

a

B

1

l!

i— I O t-H OO i— i O t— 'OOC^I I-H I-H

£

D£

•t
I

a

05 co" IM~ •*»< o o os co ^ co o~ IN" 06"

>•« t^- OS OSCJ t^-CO O^t-iCT-H T— I OS

T-H i-H C<J 1C i-l CO

CO

O^ i^H _K(

|

t-H ^OS °« CO

_°s H ^^ K

CO CO . ~ i-H CO T-H I-H CO ^^ CO *C

C

OS OS T-H *— J OS OS OS OS OS OS *~1 OS OS r~"^ ^ *"TJ
i-H i— 1 i-H •< ^-( i— Ii-Hi— 1 T-H t-l <tli— li— I<S3 •< •<

OS

5
1

^ ^ ^oboo-| g^ & ^

~ fl CO I*T\ ^H t^w -"^ CH CO CO ^ fn Cj . « CJ

(2

^— ^ 1 ^^ 1 1 HM ^ T^ | '~rf\ "*^ ^ 'TIJ *^

c3c3 < S SS S^ els^^s "<•?%.

w a
"be g
c c

O 9i

* £ 2

[136]

MUTATION IN MATTHIOLA 279

respects, and vice versa. Under favorable conditions this type may
closely resemble Snowflake, but is decidedly more slender in stems,
leaves, and pedicels. A characteristic drooping of flowers and branches
is well shown by two plants in figure 18; the single is 25b of the
tables. The progeny of 25b shown in figure 19 illustrate a variability
of the "slender" characteristics which has suggested the presence of
genetic differences among plants classed as slender. The leaves often
resemble those of crenate more closely than do Snowflake leaves.

In the field at Ithaca flowering was markedly earlier than with
Snowflake, and the type seems to be earlier on the whole. The River-
side conditions have commonly given a decided dwarfing as compared
with Snowflake, though not to the extreme degree that this has occurred
with crenate (figs. 20 and 21).

The results of selfing tests are reported in tables 34 and 35. The
distributions have the same general characteristics as with crenate,
with some remarkable differences. The excess of doubles with Snow-
flake is very much greater, the ratio being about 30 : 1 ; with slender,
however, the excess of singles is slight in the grand total and perhaps
significantly variable with different parents.

Plant 25b-ll, the "extreme" individual of figure 19, appears to
give a real excess of slender over Snowflake, and of double slender
over single slender, though the numbers are much too small for cer-
tainty. The two parents classed as ''extreme" are (tables 39 and 40 )21
quite probably genetically different from the other slender parents.
It should be noted that plant 25b-6-8-6, progeny of one of the parents
described as ' ' extreme, ' ' has also given a relatively high proportion of
slender progeny. Perhaps the "extreme" form is heterozygous for a
second slenderness factor similar to the original one.

The percentages of mutant-type progeny are (table 39) much more
variable than with smooth, large, or crenate, and (table 40) there is
no good evidence of selective elimination ; both these facts may depend
on genetic differences among the parents tested.

The great modifiability of the various types, including Snowflake,
indicated by a comparison of, for instance, figures 14, 15, and 16,
greatly complicates the positive determination of types. In the cul-
tures of 1911H and 1913, where crowing in flats or aphis injury in
the field interfered with normal development of some plants, the im-
pression was obtained that the slender type occurred in several grades

21 In the calculation of the probability of simple sampling, f is taken as 3
(the number of cultures), not 2 (the number of parents).

[1371

MISCELLANEOUS STUDIES

^Ot~ OOOO^-O -fOCO^

.,

CO ^ C^l CO ^ C^ C^ CO CO ^ (^ CO *O W

a

8

•H-HI-0 -H-H-H-H-H -H-H-H-H-H

(£

•*t>-eo«o <NOO»OOCO ^-icooo(N<y;

lO 1C ^H to O CO CO <N 1C CO COCOOCOCO
C1^ »~^ C^ **J* ^^ ^^ CO 0^ CO ^^ CO CO CO C3

i

—
a

S,

53

1

?— i O3 CO *"H ^O f» C^ CO O3 ^O C3 C^ *O

s

3

^H (N (N O 1-1 -«*i CO I-H CO (N -f (N r-i iO

"c

^1

"8

_c

1

5

i-H i-l <N CO •* •* ^ CO Tf OO i-liO 00 00 1C
t-^iCC^b- ^C^^fOOO >CcOt^-COO

|

p

1

Q

3

x

>,

-

•o

I

a

V

9

0

£

c

h

I

1

i— IOi— t i— IOO>— IC^>— 1 r- 1 C »— t i— 1 i-H r— 1

V

•d

a

t>

Oo do OQ -^ "^®t SO -^-^-

•5

c

2

1C t^ (N (N CO CO OO O (M CO 1C O --H (N 00

de

t^» CO »^ 00 CO *C G) i—l GS C3 O5 C5 *— I »•"! Oi

i-l N i-i d »C i— I d i— 1 CO i— 1 CO

3 O O
O O O

|

P- I ll

"3

°^ CO "^ sj oj

O

of c3~ of '"H'i» -S-S

000cort*ica3 S6

JJJsSSjSsSS <ao<<

(M ^^ ^Q

__! n

•

2-S-s "§1

B

g

T^^HTtiidNi-Ht^oOojajp^' "re's

<D

"2 "5 - ~

a

A

CUOCUCJCUQJCPQJfliC^ ^ O CU CU ^>

JJJJJJJJ - - «SJJ 2-S

X (M

[138]

MUTATION IN MATTHIOLA 281

probably unlike genetically. In the 1916 cultures, on the other hand,
with better development, this type seemed substantially as uniform
as the others.

If we ignore these possible genetic differences and attempt to
apply the scheme worked out for crenate, difficulties appear at once.
First, the scarcity of Snownake singles would indicate much closer
linkage than with crenate, while the relative abundance of slender
doubles apparently contradicts this supposition. Second, the in-
adequate results from crossing with Snownake (table 36) suggest
that the sperms carry the supposedly crossover slender factor at least
as often as do the eggs. While crenate as pollen parent gives results
agreeing tolerably with the hypothesis, slender gives results differing
from these in the wrong direction.

No doubt, however, the disagreements can be over emphasized. Both
crenate and slender as seed parent seem to give the expected relations
between singles and doubles, and series 23 also does this with the
Snownake progeny. Obviously the functional sperms and eggs of these
mutant-type parents exhibit different ratios between types, and the
peculiar results in other respects with slender may be related to the
added complication suggested above. The astonishing feature of the
data, of course, is the great excess of single slender over double slender
in series 23 — an excess which suggests an actual significant excess of
singles in the totals of all types given by this cross — while with selfed
slender there is a great total deficiency of singles. We may at least
feel confident that the modifications of the single-double ratio, with
this type and with crenate, are due to lethal action which also affects
the proportions of viable slender and crenate gametes or zygotes.

If differential viability before germination is an important factor
with these types, very probably it differs according as Snownake or
the mutant type is the seed parent, and according to the parental
environment. In other words, partial selective elimination during
seed formation may vary with the environment of the embryos, accord-
ing as this environment is affected by either the genetic constitution
or the external environment of the seed parent. Until such uncer-
tainties are eliminated, we are hardly justified in ruling out, for
the types discussed, the probability that regular segregation and (in
the last two cases) true linkage are concerned in these phenomena. In
fact, the definite differences in ratios between reciprocal crosses and
between at least one of the crosses and selfing encourage further
attempts at satisfactory factorial analysis.

282

MISCELLANEOUS STUDIES

05

£2

^^ ^

•5 M.

21- £.

o-»

O<N <N -^ »H I-H

_

--v S'o' ^

CD CD OS CM

T-( CO 06 1-1 CM

•

J

D
£

.2

2

^-x >O O CO

iS eojo CM^

c

o

r/j

Q

i-< CO Tjl r-( »-l

41

"Kb

CCO~

a

CO ^

o3

s*^>^*x

i-H CD (N O> O »H

COCO

"3

00 OS 1O CO •*

"8

W3 (N CM •* CO

"o

•H -H -H -ti fl

S

4fl

a

a

X

3 S

l> CO CO tO ^t|

S

<! t.

&

o i-* CM I-H cs r-~

£

>.

k

_a
•fi

it

§1

5
^j

c
*

Iz;"0

CO to CO to CM CM

S

S

CM CM

£

IV

3

^TcM*

g

s ^v ^

O

Q

rt i-HO 0

"M

CM

a

CM CM

02

1

1

11

cocoeocM -H t^-

CMO3C-1CM CM

j— (

c!

X

5

,

"c3
|

1*2

«J C

T-H O O •-* O t-H

H

11

PH

f

•* iooc3sco r-

.-H Cl (N 1C tO CD

02

_l ^ CM ^H

a

•

CO CO CO • —

^

T I"1—. "T **

O

rH •— ( i— ( <J i— I ' 5 t-
OS O5 OS OS g ®

r-H t-H r-^ • ' Q^ J^

O "

a

i
•

•

=a 8 g

•n *

J3 CO
^ tn CM

£

^ «*i 00

03 o3 03 ° oJ.SJ

CO CO CO 'Tl ^t1 CJ
CM CM CM <3 <T1 CO

•a

ao
.3

0>

S.IS

a 2*
s —

"

^S ®

J 3 12

tc

'C «

5 ^a

<n _ fl

d -2 5

<D
. 01

— ^

-3 « S
- - a>
A
» +»

S «H

® OT G
OP -4J <J>

[1401

MUTATION IN MATTHIOLA

283

6. THE NAREOW-LEAVED TYPE

As table 37 indicates, this type competes with crenate for second
place in frequency of occurrence in the Ithaca cultures ; in fact, when
only the strictly unselected cultures are considered the percentage is
very close to that for smooth-leaved. A feature of special interest is
the apparent association of the mutant type with doubleness.

In a cool greenhouse this type (fig. 22) varied from exceptionally
late and many-noded to ordinary in both characters. The leaves (see
also fig. 18) were typically narrow, rather strictly entire, often rolled
backward or twisted, and typically more ascending than those of

TABLE 37

Narrow-leaved type. Numbers of apparent mutants and association of the type
with doubleness of flowers.

Culture

Progeny of Snowflake and early parents

Total
examined11

Narrow-leaved

Single

Double

All

Coefficient of
mutation

1908
1910
191 IF, house-sown
Al! above
All unseleeted

725b

338
2072
3135
2410

0

1

7
8
8

2
4
12
18
16

2
6
20
28
26

.28 ± .26
1.78 * .38
.97 ± .15
.89 * .12
1.08 =*= .14

a See note b to table 2.
b See note c to table 1.

Snowflake. The apex of the leaf is often more acute than with Snow-
flake, and many leaves are mucronate or at least end in a sharp, rigid
tip.

A striking characteristic is the narrowness of the sepals, resulting
in frequent early separation at the edges, partially exposing the petals
in immature buds.

Under the less favorable field conditions the plants often remain
long as dwarf rosettes, and flower late and feebly if at all. Figures 23
and 24 show comparatively well developed plants in the field.

The type is on the whole very distinct in the field, though there
has been some question whether a greenhouse plant such as that in
figure 18 is genetically different from those with short and rigid
leaves (figs. 22 and 24) ; the very great variability in leaf form due
to external conditions makes such a question very difficult without
extensive progeny tests. It is now (1918) probable that narrow-dark
(p. 143) was not distinguished from narrow in the greenhouse.

[141]

MISCELLANEOUS STUDIES

o 2

u ^^

°*

•-I iH N O CO

_.

Sid- S3 Si 2

CO t-l t~- r-> If.

1

•
3

o" £" o?

— — —

5

3

=
do

Q

occo -^ ^ «c

01

O x-s CO •— ' CS

"5

I— 1 CO »H 1-H f— t

d

N.^SB. S **^S *^.' +*.*'

en

00"3 CO O O
1-1 1-1

"3

-# % al

*

CO >O C4

o

c

-H -H -H

9

- S

CC i— 1 i— 1

_2

TjJ t«

•< «

O ^ Ci

i

Ej

p ^

c

e
3

5

1C <M

S
£

*

2

x — ^ o-*
r-IN CO JO, i-H

o

p

co

«
"3
c

O <-H r-l O fr-
rH

in

"8

j|

3

'S.S

C^l ^H CO C~J ^^

1

QS

X

£

,

?

^2 c

OC O O iO

•O-g

Seeds"

"t 0 00

— i~ CO CO O

co co

Oi O5 r- 1

^

=a °a ^

O

K 5 ffi °^

i-H CO •< i-1 CO
I— 1 1— 1 1— 1 1— 1

O O Ol O
I— 1 1— I — —

i 1

a

!?i~il

1

IjjS'j

Is5 ?""5I

MUTATION IN MATTHIOLA 285

The few singles have produced few seeds, and these were highly
variable in size. The capsule often has a defective septum, more or less
of the distal portion being absent. Germination was poor in the small
cultures secured (table 38, upper part), with only 10.8 per cent of the
mutant type among the progeny.

This case agrees in most respects with those previously discussed,
but adds one point of interest in the occurrence of apparent coupling
of mutant type with doubleness rather than singleness. Seed appears
to be less abundant and less well developed than with any of the pre-
ceding mutant types, facts probably significant in relation to the low
percentage of narrow progeny from narrow parents, though the large
probable error of the percentage must be considered.

7. MISCELLANEOUS ABERRANT TYPES

As part of the aberrant individuals occurring in the greenhouse
were either doubles or singles that produced no seed, while practically
no seed was produced by any plants in the field at Ithaca or by even
some of the commoner mutant types at Riverside, the opportunity for
progeny tests has been almost entirely limited to the types so far
discussed.

The narrow-dark-leaved type (table 3) was common and distinct
in the field at Ithaca, where it constituted about .48 per cent of the
2072 plants from house-sown seed, and has been readily identified
in several cases at Riverside. It was not distinguished in the green-
house cultures, but was very probably included under narrow-leaved.
Possibly a single described as "small-convex-leaved" belonged to this
type, though two field plants were given this name as distinct from
narrow-dark; according to a photograph (fig. 25, second plant from
left), another greenhouse plant (a double) may have been similar to
narrow-dark-leaved. The narrow-dark-leaved type (figs. 26 and 27)
has narrow dark-green leaves, strongly convex upward, and evidently
tends to compactness of growth and lateness of flowering ; under field
conditions it seems decidedly more like Snowflake than like narrow-
leaved.

The 44 progeny (table 38) secured from the greenhouse single
mentionad above included 2 (4) narrow-dark-leaved individuals and
3 (5) other plants not Snowflake (the last including two smooth, one
large, one slender, and one semicrenate), besides five undetermined
plants. Plainly the type of the parent is still in doubt.

286 MISCELLANEOUS STUDIES

Another very different greenhouse plant, described as "stout
dwarf" (fig. 25, third from left), gave among 29 progeny (table 38)
5 (7) individuals evidently not Snowflake, which may have been
narrow-dark or may have belonged to another type that was somewhat
similar under the conditions of the tests. The parent resembled
Snowflake except in its short internodes and short, stout capsules.

Four other plants suspected of mutation apparently entirely failed
to repeat their type in their progeny, perhaps because of the smallness
of the house cultures. One of these was the plant, much branched
for the warm greenhouse, third from the right in figure 18; another
was a very late plant with a remarkably large number of main-stem
leaves; the others were a plant with unusually small flowers and one
with some of the leaves somewhat spatulate. Possibly all of these were
Snowflake, though the second, which gave poor germination, probably
was not. All these four plants have been included as Snowflake
parents for tables showing numbers of apparent mutants.

The small-smooth-leaved type is well shown in figure 25 (first and
fifth from the left) . It is the smallest and weakest of the fairly common
and definitely identified types ; it has small, very smooth leaves, and is
late in blooming. The two plants shown were both singles, but they
set no seed.

The semicrenate-leaved type (table 3) differed slightly but appar-
ently definitely from Snowflake, somewhat resembling crenate-leaved
in leaf form. The one "pointed-crenate-leaved" plant of table 3 may
have been crenate-leaved. The "compact" and "curly-leaved" plants
of this table have not been identified with any aberrant types in other
cultures. With the remaining six types of table 3 all the individuals
have been questioned as possibly Snowflake ; it is now practically cer-
tain that some of those in the second, third, and fourth groups
belonged to the large-leaved type since studied, but the apparent inter-
gradation with Snowflake makes any attempt at a definite reclassi-
fication from the records a matter of doubtful value.

The second plant from the right in figure 25 was remarkable for
its short stem and few but large leaves. Several other more or less
exceptional individuals have appeared in the cultures, especially among
some plants with abnormal cotyledons, selected from large numbers of
greenhouse seedlings in the 1908 cultures, \vhich were examined for
syncotyledony. Some of these were very weak plants which finally
died without flowering.

f!441

MUTATION IN MATTHIOLA 287

The fluctuations in habit, leaf form, etc., within the type are such
that the determination of familiar types is often a matter of some
uncertainty, as is shown by data that have been presented. It may
well be that among the doubtful types are included several definite but
comparative rare mutant forms, which occurred too infrequently to
afford adequate material for positive classification.

8. SOME PROBABILITES OF EANDOM SAMPLING

For compactness of presentation and convenience of comparison
the material in tables 39 and 40, to which some incidental references
have already been made, is collected here rather than scattered through
the discussions of the various types concerned. Some statements as
to methods are also necessary in connection with each of the topics
here treated.

First, it should be noted that the percentages previously given
have regularly been accompanied by the probable errors of simple
sampling. These probable errors have been calculated by the formula

per cent = .6744898 ^L , where p is the percentage of the mutant

n
type ("successes"), q is 1 — p, and n is the size of the sample (the

number of plants concerned).

In the heredity tables for each type, p has uniformly been taken
as the percentage of the total of the lots compared, or p0.

For the "mutation coefficient" the percentage of the grand total
of unselected house-sown lots has regularly been used. Evidently the
few selected progeny included in tables 1, 28, and 37 should be omitted.
All the percentages here are so low that the probable errors deserve
little confidence, even though n is usually fairly large. The rather
close agreement of the percentages of all apparent mutants in the
three distinct lots of unselected house-sown cultures suggests that
they represent fairly well the population value for the potentialities
of the seeds ; and even if the mean percentage of the total of the lots
for the main comparisons is actually nearer, it is safer to use the
larger probable errors resulting from the method here employed.
Furthermore strict use of p0 would sometimes require several slightly
different probable errors for the same percentage, for use in different
comparisons in the same table.

[145]

288

MISCELLANEOUS STUDIES

If the probable error of the difference of any two percentages in
the same table is to be obtained, therefore, formulae corresponding to
those given by Yule (1911, pp. 264-267) are applicable.

Now, it is possible in some of these cases to calculate the actual
standard deviation of the percentage in subsamples which make up
an aggregate sample. Table 39 gives such actual standard deviations,
in comparison with the corresponding theoretical or expected standard
deviations given by ,

_ y pq

" per cent I— •

n — 3

TABLE 39

Standard deviations of percentages of mutant types. Values derived from V pq,
compared with values expressing the actual variability of subsamples.

Type of parent and grouping of
progeny

N

'

n

p

Standard deviation of samples of mean size n

Actual

Theoretical.-J p 9

Difference

n-3

Eff

Smooth-leaved type:

All lots by parentage

234

6

39.0

27.35

7.5

7.4 * 1.4

+ -1

All lots as grown

234

12

19.5

27.35

11.3

11.0 * 1.5

+ -2

Germination good

187

7

26.7

29.95

10.9

f 9.4 ± 1.7

\ 9.2»

+ .9

Germination poor

47

5

9.4

17.02

5.2

(14.9 ± 3.2
\17.6

- 3.0

Large-leaved type:

All lots by parentage

357

20

17.85

49.02

10.7

13.0 ± 1.4

- 1.6

All lots as grown

357

22

16.2

49.02

10.9

13.7 ± 1.4

- 2.0

Germination good

260

14

18.6

50.38

11.3

f 12.7 ± 1.6
I 12.7

g

Germination poor

97

8

12.1

45.36

8.7

/16.5 ± 2.8
\16.6

- 2.8

Crenate-leaved type:

All lots by parentage

633

20

31.65

29.86

10.6

8.6 ± .9

+ 2.2

All lots as grown

633

28

22.6

29.86

12.5

10.3 ± .9

+ 2.4

Germination good

549

20

27.45

32.42

10.7

f 9.5 ± 1.0
I 9.3

+ 1-2

Germination poor

84

8

10.5

13.10

10.5

f 12.3 ±2.1
\16.7

- .9

Seed-size test, smaller seeds

73

5

14.6

65.75

13.2

/13.9 ± 3.0
1 13 8

- .2

Same, larger seeds

193

5

38.6

20.21

9.4

f 6.7 ± 1.4

{ 7.9

+ 1.9

Same, all seeds, by parentage

266

5

53.2

32.71

10.3

6.6 ± 1.4

+ 2.6

Same, all seeds, as grown

266

10

26.6

32.71

22.9

9.7 ="= 1.5

+ 8.8

Slender type:

All lots by parentage

243

8

30.4

32.51

17.5

9.0 ± 1.5

+ 5.7

All lots as grown

243

13

18.7

32.51

19.7

11.8 ± 1.6

+ 4.9

Germination good

165

7

23.6

33.33

14.9

f 10.4 ± 1.9
\10.3

+ 2.4

Germination poor

78

6

13 0

30.77

27.2

f 14.6 ± 2.8
114.8

+ 4.5

Parents "extreme"

38

3

12 7

63.16

14.4

f 15.5 ± 4.3
115.1

- .3

Parents ' ' ordinary ' '

205

10

20.5

26.83

14.7

( 10.6 ± 1.6
111.2

+ 2.6

Narrow-leaved type:

All lots as grown

37

3

12.3

10.81

8.1

10.2 ± 2.8

- .75

"The second values for some cases in this column are derived from pa (see text).

[1461

MUTATION IN MATTHIOLA 289

For example, table 27 gives the percentage of large-leaved plants

among the 357 progeny of the 20 large-leaved parents as 49.0 ± 1.8

i —

per cent. This probable error is given by .6744898 >— , where

n

p = 49.0 per cent, g = 51.0 per cent, and n= 357. These 357
progeny, as table 39 indicates, came from 20 parents which contributed
an average of 17.85 progeny each, and the actual standard deviation
of the percentage in these 20 sibships was 10.7 per cent.

Obviously the expected standard deviation of simple sampling for
comparison must represent samples not of 357 plants each but of 17.85
plants each. Now a percentage is obviously a mean (of values all
either 0 or 1). Since "Student" (1908) has shown that the theoretical
standard deviation of the mean in samples is given more exactly by

°°varlate ,-1 -i O" variate

<T

mean

,-1 -i varat

than by <rmean = - =

'-M 9

ft O

(the value for the normal curve conventionally used for the probable
error of the mean) and since w, the mean size of sample, is small
enough to make the correction a matter of considerable importance.
— 3 is here used. Since o-variate = VP<?> we have o-mean =

, where >n = 17.85. This gives a theoretical standard devia-
tion of 13.0 per cent.22

It is true (Yule, 1911, p. 260) that the ordinary method of calcu-
lation of the actual standard deviation is not satisfactory for means
when the samples vary in size. A method has been used, however, which
obviates this difficulty, so that comparison with the results given by

j - Q ig strictly legitimate. Each squared percentage deviation

71 o

has been weighted by multiplying it by the number of individual
plants which it represents, and the summation of squared deviations
has then been divided, not by 2/, the number of samples, but by
2/ X w, the number of samples multiplied by the mean weight or
average size of sample (in other words, by N, the total number of
individuals).23

22 In the calculations for table 39 p has been taken as the percentage given
in this table, to two decimal places, while with all other numbers employed in
calculation, including n — 3, three or more decimal places have been used as
needed.

23 Algebraic proof of the correctness of the method has kindly been furnished
by Frank L. Griffin, Professor of Mathematics, Reed College, Portland, Oregon.
If it develops that this rather obvious device has not been suggested for the
purpose, it is to be presented elsewhere with the mathematical proof. When the
variates are not grouped in classes the calculation is substantially as easy as
without weighting, while the theoretical value is found with much less work
than by the method given by Yule (1911, p. 260), which requires the harmonic
mean of the sample sizes.

[147]

290

MISCELLANEOUS STUDIES

II

« o

SJ «*

s o

to '«?

c

I

a

1

^o

.0

- g

•2

1—1

•

t

C^ Is* I-H *C CO CO

0 ^

•3
||

1
1

00<N COO CSi-n-i CO
*O ^t1 O^ Ot»
.-1 M CO O

1

N^ O

l-H i-H

i

9}

sa

•33

i-H

•D

0."°

O

o o.

c

M g

^0

•2 41

'•3

Oi

s a

'i

O CO

'•-I

«•> Cv

•3

CD i-H
Tt< 00

to

•3

•s

CO1*? ICCC 00»CQ O

O

t^ O

.2

"a

o
o

a

<-l(N (NO OCOO O

§

1C O

0 0

S

06

a

—

I

a

O

1

o

0

£

'j

US

— ^

^~.

1

o

c

B

c

OC 00 O (N IN CO

O CO

ll

a

C3

(N »-" O »C
CO "3
fH

CD

•-H

00 •<*

1C

1|

•3

s

|f

c

O

a

is

1

03

•c

|

"o

CD CO O O •* (N
-* CO 1-1 IN CO 00

CO

22 §0

1

1

CD Tf Oi CO O O
C<l 1C f O O 1^-

8

l-H i— 1
^H O

6

C

.

•

iL

aj |r,

"S « r^ « Is-* aj |^

o

•S « r^

s

.S ^j & h~> ^[s~> >! h~*

cc ^

"S ws~'

—

S "^ ^"^^ "** ^> •*•* 03 ^^

"S **

o *» ^\

3

i-1 fH ^> T £} ^ j^ O T

S^ 0}

" u ^

o

£

bc"o a> II g II T3 _>> II

^iJs ^c "^•a'3 ^

SI

is

9

02 ^ II
> J^

5g| || ||.| |

£.|

111 *

1
o

"S

CO

I1

no -^

<C T3

be
c 13

'» 3

3 -s

»r cs

[148]

MUTATION IN MATTHIOLA 291

In the calculation the deviations are taken from zero, and with
these small numbers of samples the percentages are not thrown into
classes; it suffices, then, to square each number of "successes," divide
by the corresponding total of individuals, add the quotients, and
divide by the grand total of individuals, correcting this weighted
mean squared deviation by subtracting the square of the weighted
mean percentage (percentage of grand total). If s is the number
of successes and n is the total number of individuals in the sub-

2s
sample, and M is the weighted mean percentage, then M =—— , and

s-

ent— I ^W M

Table 39 gives, for the most important comparisons of heredity
percentages, the total number of progeny (N)> the number of cultural
groups or (with the first line for each type) the number of parents (/),
the average size of the groups of progeny (M), and the mean per-
centage of the mutant type (p). This serves as a summary of some
of the most important statistical data already presented relating to
the inheritance of these types, and also shows the basis of the remain-
ing part of this table and of table .40. For comparison of actual and
theoretical standard deviations the theoretical value has been calculated
from the actual percentage as given in this table. For comparison of
means (table 40) the percentage of the corresponding total (p0) has
also been used, this theoretical standard deviation being the second in
the table in the cases where the two values are not identical.

Since small changes in a percentage have little effect on its
theoretical standard deviation, we are fairly well justified in taking
the latter, as calculated from the actual percentage in each case, to be
the ''population" value. Consequently, the difference between the
theoretical and actual standard deviations has been expressed in each
case as a multiple of the probable error of the theoretical value.

Aside from the last line for crenate-leaved, where there is an
obvious artificial reason for high variability, there is no very significant
difference except with slender. In this case, the deviation of 5.7 times
the probable error (line 1) is probably largely due to the genetic
differentiation of "extreme" and "ordinary" parents suggested by
their appearance and by the wide difference in the heredity per-
centages; the differences become moderate when the progeny of the
two classes of parents are separated.

[149]

292 MISCELLANEOUS STUDIES

In the two cases (smooth-leaved and crenate-leaved types) where
the percentages of mutant types differ greatly with good and poor
germination, separation according to germination gives a mean value
of the standard deviation decidedly lower than the value for all lots
taken together. In the case of the large-leaved type there is little
change, while the considerable reduction with the slender type is
probably due to unequal separation of lots from parents genetically
different.

Table 40 shows the simple-sampling probability of the most striking
differences of heredity percentages, aside from the characteristic
differences between different types. "Student's" (1917) table of
probabilities of mean deviations with small samples is used, with
interpolation by second differences. Where the standard deviation
of the difference is required it is found from the theoretical values
given in table 39 by the formula (Yule, 1911, pp. 264-265)

— -»/ I _1^ ^0 ^0 I
<T difference V °"l ~T~ °2 " — 5"T

Po Qo , Po Qo

O I O

when one statistical population is assumed (table 40, columns 2 and 3).
When two populations are assumed (table 40, columns 4 and 5) the cor-
responding formula using plq1 and p2q2 is employed. In the one case
where this is possible (the seed-size test), it is also calculated from the
actual differences of the pairs of percentages in the separate tests, each
difference being weighted with the total number of progeny from the
parent concerned. Where two values of / (the n of "Student's"
table) are involved, the smaller is taken, giving understatements of the
probabilities involved; in the two cases where the difference is more
than 2, the values are recalculated, with / as the nearest smaller
integer to the geometric mean of the two actual numbers (that is
with /„= V/i/2)- 1° the case where the probabilities of four devia-
tions all in the same direction are combined, the four chances of
occurrence are multiplied together; that is, if the £(l-f-a) of
"Student's" table is P, and l — P is F, then Fl.2.3.t = F^F^-F^F^
"Student" (1908, p. 1) says, "The usual method of determining
the probability that the mean of the population lies within a given
distance of the mean of the sample, is to assume a normal distribution
about the mean of the sample . . . ." When this is done with a differ-
ence of means, it is at once evident that only half of the chances of
deviations as great as the distance of the given difference from zero
difference lie below zero difference; the other half of the chances of

f!50]

MUTATION IN MATTHIOLA 293

such deviations lie in the opposite direction and represent positive
differences still greater than the sample difference. In other words,
if the implications of a sample difference are to be given full weight,
this difference must be considered the most probable value of the
theoretical "true" difference between two assumed distinct statistical
populations. In the present case we wish to know the probability that
the "true" or theoretical-population means differ in the same sense
as the observed sample means. This involves calculation of the proba-
bility of deviations in one direction (beyond zero difference) from
the sample difference. If the sample difference of means is considered
as positive, then the negative "tail" of the theoretical frequency
curve of sample differences (this curve being centered at the observed
sample difference) must be compared with the rest of the curve. The
positive portion of the curve the ^ (1 -|- a)24 of the tables, then gives
the chances favoring the hypothesis that the sample means truly
represent the population means. The odds in favor of the hypothesis
are therefore given by the formula

or

Values calculated from this formula are given in columns 4 and 5 of
table 40.

When other considerations than the sample evidence are to be taken
as determining the most probable value of the "true" mean, the case
is different. For example, if the probability that our sample per-
centages are mere sampling deviations from some theoretical Mendelian
value were in question, that theoretical value must be taken as the
population mean and only the magnitude of the deviations must be
considered.

When a difference of means is considered from this latter stand-
point, it is assumed that the two samples come from one statistical
population, and hence that zero is the most probable value of the
population difference. If we choose to assume that the most probable
value of the, population difference in our cases is zero, we must
calculate the odds against a deviation of the observed amount in
either direction from zero difference. The formula for these odds is

°2 = 9 VIM T~ ' or T~^ ~~ '

* s\ a \+ — a) * — a

24 The whole area of the frequency curve is taken as unity, and a is the area
enclosed by any given deviation in both directions from the mean.

[151]

294 MISCELLANEOUS STUDIES

Values from this formula are given in columns 2 and 3 of table 40 ;
their magnitude in three cases, however, and the uniform agreement
of the direction of difference with the expectation from biological
evidence which has been discussed, weigh heavily in each test against
the assumption of random sampling from a single statistical population.

It does not appear necessary, however, thus to weigh the evidence
in detail before deciding which formula is suited to the case. There
is no evident theoretical value from which these percentages are
reasonably likely to be sampling deviations. This being the case, and
granting such general possibilities as that of differential viability, it
seems most reasonable to use the former (OJ formula. That is, we
should give full weight to the implications of a sample deviation
unless there is some definite reason for assuming that some other value
better represents the mean of the theoretical statistical population.

It must be remembered that the actual probabilities of sampling
deviations do not necessarily correspond closely with the probabilities
of random sampling. With the material in table 40, however, aside
from the germination comparison in the case of the slender type,
table 39 suggests a fair agreement with the conditions of random
sampling. The actual standard deviations of the subsamples do not
in general differ widely from the corresponding theoretical values, and
the differences are negative about as often as positive.

The hypothesis of selective elimination with poor germination is
strongly sustained (table 40), although only one difference (with the
crenate type) has much statistical significance when considered alone.
If we may multiply together the members of the four ratios in column
3 of the table, the combined odds (using the /0 values) are 130:1
against occurrence of these four deviations as accidents of simple
sampling, when magnitude of deviation alone is considered. If
direction of deviation alone is considered the random chance of these
four deviations all in the same direction is obviously (|)4, or the odds
favoring the elimination hypothesis are 15:1. Combination, of these
two chances indicates a high probability for the hypothesis. When
the two-population formula is used in calculating the standard devia-
tion of the difference (columns 4 and 5) the value of P is consider-
ably reduced in some cases, and the combined odds obtained from
-Fx • F2 • -F3 • F± are very high. Evidently the best single expression
of the simple-sampling odds, though possibly somewhat too high, is the
value given last in column 5, or 123,093:1.

With the seed-size test of crenate the odds are 499 : 1 with the
theoretical standard deviation of the difference, or 1666 : 1 with the

[152]

295

actual standard deviation. When the relatively small size and weak
growth of crenate seedlings are also taken into account, the relatively
small average size of crenate embryos may be considered to be
demonstrated beyond reasonable doubt.

With "extreme" and "ordinary" slender parents the odds de-
cidedly favor the hypothesis of genetic differentiation of parents, in
spite of the small numbers involved. We must remember that definite
statistical differentiation of lots of progeny grown under uniform con-
ditions does not necessarily demonstrate genetic differences (differences
in output of gametes) between the parents; in this case, however, the
difference in the appearance of the parents and in the single-double
ratio among the progeny also suggest genetic differentiation.

GENERAL DISCUSSION25

It might be argued with some plausibility that the available
evidence hardly justifies conventional factorial analysis, or at least that
the data indicate strongly the presence of marked factorial incon-
stancy. The aberrant types occur in very small proportions among
the progeny of selfed Snowflake parents, in much larger proportions
from "mutant-type" parents, and in intermediate proportions from
crosses with Snowflake. It might be supposed that the Snowflake type
has a slight tendency to mutate to the other types, and that these have
a much more marked tendency to mutate back to Snowflake. Various
considerations, however, especially the occurrence of apparently
regular linkage phenomena, seem to favor the general form of
hypothesis which has been presented.

As we have seen, it is well known from the behavior of various
factors that the typical Mendelian mechanism is present in Matthiola.
It cannot be argued here, as sometimes with Oenothera, that the
genetic behavior of the genus or species is fundamentally non-
Mendelian. Since the Mendelian mechanism is demonstrably present,
and Muller's (1918) work on beaded wings in Drosopkila seems to
establish the adequacy of this 'mechanism in a closely parallel case,
surely conventional factorial analysis should be carried as far as pos-
sible; in fact (Muller, 1918, p. 423), a Mendelian explanation should
not be abandoned for anything short of positively contradictory
evidence.

2.r> Muller's (1918) complete report on the beaded-wing case in Drosophila
appeared several months after the present paper had gone to the publisher.
Certain conclusions given below, very similar to Muller's but not credited to
him, were therefore reached independently.

[153]

296 MISCELLANEOUS STUDIES

In the Drosophila case just mentioned, the "principal" factor for
the character in question is "dominant for its visible effect and
recessive for a lethal effect," so that no pure beaded individuals
appear among the progeny of beaded. The original race regularly
gave progeny partly heterozygous beaded and partly homozygous
normal, while after a long period of selection a true-breeding beaded
race appeared. This latter form, it proved, fails to give normals not
because of being duplex for beaded — it is still simplex — but because
of its possession of another factor, known only by its lethal effect
when homozygous, which is carried by the chromosome bearing the
normal allelomorph of the factor for beaded. The locus of this reces-
sive lethal factor gives in general about 10 per cent of crossovers with
the locus of beaded, but in this case, because of the presence of a factor
"which almost entirely prevents crossing over" between the loci of
the two lethal factors, viable non-beaded zygotes are very rarely
produced. Thus every zygote receiving either two beaded-carrying
chromosomes or two non-beaded-carrying chromosomes of the pair
concerned fails to develop, and all the insects produced are necessarily
heterozygous for both lethal factors.

A point of special interest in this case is the fact that by certain
crosses individuals can be produced which give certain types among
their progeny in very small percentages. Muller suggests that part
at least of the supposed mutants of Oenothera may be due to crossing
over between chromosomes carrying lethal factors, by which certain
recessive factors are permitted to come to expression in viable zygotes.

For the inheritance of doubleness of flowers in Matthiola he gives
a "balanced-factor" explanation essentially identical with mine (Frost,
1915).

There seems to be little reason to doubt that the differential factors
for these aberrant Matthiola types have originated by mutation. On
the analogy of Drosophila we might expect that the true mutations
would be relatively rare, and that most of the apparent mutants, in
cases where they appear frequently, would be due to segregation,
appearing as the result of crossing over in chromosomes carrying
balanced lethal factors. The evidence seems to indicate, however, that
the differential factors for the mutant types at all extensively studied
are dominant for their visible effects and usually (probably imper-
fectly) recessive for a lethal effect, the mutant factors thus being
genetically similar to the factor for beaded wings in Drosophila.
This would seem to imply the occurrence of certain mutations in pro-

[1541

MUTATION IN MATTHIOLA 297

portions as high as about 1 per cent, and a general mutation coefficient
of perhaps 4.5 per cent, while the only Mendelian alternative would
seem to be some more complex scheme whose satisfactory formulation
might require much more extensive hybridization data.

To be more specific : ( 1 ) these types are not single recessives, since
they are not homozygous but split into the mutant and "normal"
types; (2) they are not simple cases of multiple recessives, as has
been proposed by Heribert-Nilsson (1915) for Oenothera mutations,
since what is on that hypothesis the full dominant type reappears with
selfing; (3) if these types are single dominants, as they appear to be,
they cannot (barring the action of inhibiting factors) arise from the
pure recessive "normal" or Snowflake type by segregation, but only
by immediate mutation; (4) they are not simple cases of comple-
mentary dominant factors, since they occur among the progeny of
selfed parents.

We might assume that a "mutant" type depends on two pairs of
factors, one homozygous and the other heterozygous, while both pairs
are heterozygous in the "mutating" Snowflake parent. Thus the

D Ci

crenate type might have the zygotic formula -= J— , where d is the

d ci

factor for double flowers, C a dominant factor for crenate, and / a
dominant inhibitor of C, all three loci being situated in the same
chromosome, at distances of, say, 16 and 4 units apart, in the order
indicated. A Snowflake parent producing crenate progeny would

then be -5 — - or -^ — - , and crossover combinations would produce the
dci del

apparently mutant crenate progeny. The crenate progeny would
behave as heterozygous dominants when selfed, and if CC zygotes
were non-viable would yield constant Snowflake and inconstant
crenate; the extracted Snowflake singles, having the composition

Dci

-T— r, could not throw crenate individuals except by true mutation of

c to C. With selfed Snowflake, if we assume 16 per cent and 4 per
cent of crossing over in the two positions, and a 60-per-cent selective
elimination of crenate zygotes, all CC zygotes being non-viable, sub-
stantially the observed percentages of crenate singles and doubles
result.26

20 See page 125, footnote. This scheme agrees fairly well with the results
from crossing, and gives almost exactly the observed proportion of total ' doubles
(a little over 53 per cent) for selfed Snowflake. Its adequate presentation must
be reserved for a later paper.

[155]

298 MISCELLANEOUS STUDIES

Formerly (Frost, 1916) the hypothesis of frequent dominant
mutations seemed the more probable, but there is apparently non-
conformable evidence. It is true that the peculiar behavior of the
slender type might conceivably depend on an occasional mutation in
another locus, or an exchange (Shull, 1914) or duplication of loci,
giving two similar or identical factors for slender. An apparently
fatal objection, however, is the fact that the supposed mutants seem
to show linkage with singleness or doubleness at their origin from
Snowflake as well as in subsequent generations — a fact which strongly
suggests segregation in the former case.

If the apparent mutants are really due to segregation complicated
by lethal action, the origin of the complex heterozygosis indicated for
Snowflake is doubtful; it may be due to hybridization, but more
probably to a gradual accumulation of mutant factors in balanced-
lethal combinations. On the analogy of Muller 's Drosophila case,
especially, it might be expected that the latter would be the true
explanation, particularly since self fertilization seems to be the rule
in Matthiola. On this basis the term mutant type is used with some
confidence in this paper, while the aberrant individuals have been
called apparent mutants.

We must not forget that some of the mutant types may arise, as
with Oenothera gigas and 0. lata, by non-disjunction, or reduplication
of chromosomes, and that this fact may determine their heredity.
This is not to be expected with the types whose factors show apparent
coupling with singleness or doubleness, but it might be true of the
apparently unlinked smooth-leaved type. A preliminary study of
several types show's that the usual somatic number of chromosomes
is probably fourteen, but that positive counts are difficult. While it
might be very hard to demonstrate the regular presence of one extra
chromosome in an individual or a type, it should be easy to decide
between the diploid and triploid numbers. The large-leaved type is
so strongly suggestive of 0. gigas that it would not be surprising to
find the triploid number in the material now on hand for examination.

In a preliminary paper on these types the writer (Frost, 1916)
discussed some possible relations of mutation, heterozygosis, and
partial sterility, with special reference to Oenothera, mentioning the
possibility that special prevalence of heterozygosis in the genus may
be, "in large part, a result rather than a cause of mutation." This
suggestion is evidently justified even if much of the supposed mutation
of Oenothera is really segregation, since it is highly probable that

f!56]

MUTATION IN MATTHIOLA

299

the peculiar phenomena depend on lethal factors or combinations of
factors originally due to mutation.

Another possibility there mentioned, advanced by Belling (1914)
and since specially discussed by Goodspeed and Clausen (1917), is
that of the occurrence of lethal combinations of certain factors which
in other combinations may be in no way prejudicial to normal develop-
ment. As the latter paper shows, it is probable that in certain
crosses between ' ' good species ' ' most of the new combinations brought
together in the formation of Px gametes are incompatible with the
production of functional gametes. Perhaps in the case of Oenothera
there may exist within a species factors lethal in any combination
when homozygous, and other factors lethal only in certain com-
binations.

A balanced-factor explanation for the inheritance of doubleness27
in Matthiola, a case which Muller (1918) discusses, seems to have been
first definitely stated by Goldschmidt (1913), though he failed to pro-
vide for one feature of the evidence, the deviation of the heredity
ratio from 50 per cent. As has been shown (Frost, 1915), this
peculiarity may be due to greater viability of the homozygotes (sterile
doubles) during embryonic development, since the doubles are more
viable in the mature seeds and more vigorous in later development
(Saunders, 1915). In this case the "normal" factor is completely
eliminated in favor of the mutant (sterile-double) factor in the
formation of the sperms, and probably is partially eliminated in the
formation of either the eggs or the embryos or both.

Here the normal singleness (sporophyll) factor D may act as a
lethal in the heterozygous parent, possibly from its general relations
of growth vigor in the presence of the more vigorous d-carrying cells.
If the lethal factor is situated in a distinct locus, evidently crossovers
are at most extremely rare. It is true that Miss Saunders (1911)
finds that Fj hybrids with pure single forms produce functional
single-carrying pollen ; with the pure single forms from which the
original "double-throwing" mutants arose, however, this might not
be true, or a lethal change may have occurred in the singleness factor
itself rather than in a factor coupled with it. The Drosophila case
would suggest a lethal change in another locus of the single-carrying
chromosome.

In my paper of 1915 this lethal change in one chromosome ap-
parently accompanying the mutation of D to d in the homologous

27 For a brief outline of the genetic behavior of doubleness see the discussion
of the experimental data for the smooth-leaved type.

[157]

300 MISCELLANEOUS STUDIES

chromosome was considered puzzling. Evidently, however, it may
have occurred in one chromosome before D mutated to d in the other,
and even then may have produced its lethal effect. It is evident
that if doubleness should arise in the absence of the lethal effect it
would tend to be eliminated by the return of one-third of the singles
to the homozygous condition in each generation. In fact, it is possible
that the lethal change arose later than doubleness, as in the Droso-
phila case, or was brought in later by cross pollination, and happened
to be preserved as a result of horticultural selection for a high pro-
portion of doubles.

A parallel-column comparison between the double type and the
types especially discussed above has already been given, in connec-
tion with the smooth-leaved type. It will now be seen that this com-
parison seems to apply to all mutant types, except early, that have
been genetically tested, the principal differences between these types
relating to the heredity percentage and the apparent presence or
absence of linkage with the single-double factors.

From the standpoint of its relation to genetic analysis the double-
ness factor is remarkably similar to the sex factor in animals. There
are two types in each generation, one heterozygous and the other
evidently homozygous, and these types are produced by the fertiliza-
tion of two kinds of eggs, produced in equal or nearly equal numbers,
by a single kind of sperm. Although one of the somatic types is
sterile, and the uniformity of the sperms produced by the other is due
(evidently) to lethal action, the opportunity for chromosome analysis
is similar to that with sex chromosomes.

We may say that the doubleness factor and its normal allelomorph
(d and D) are carried by chromosome pair I. Already we know
several other pairs of factors evidently carried by this pair of chromo-
somes. These are, to name only the mutant or possibly mutant
member of each pair of factors: P (pale sap color) and W (colorless
plastids), both studied by Miss Saunders (1911, 1911a) ; C (crenate-
leaved), S (slender; possibly two factors), and N (narrow-leaved).
As we have seen, the last three of these are probably lethal when
homozygous, and one or more unidentified lethal factors may be con-
cerned in the breeding results, while the doubleness factor affects the
race much like a recessive lethal, since all dd individuals are completely
sterile.

[158]

MUTATION IN MATTHIOLA 301

SUMMARY

This paper describes the occurrence, characteristics, and heredity
of certain aberrant plant types which decidedly resemble some of the
" mutant" types produced by Oenothera lamarckiana. The parent
form is Matthiola annua Sweet, of the horticultural variety "Snow-
flake."

These aberrant forms may be called mutant types, since it is highly
probable that they are originally produced by mutation. The aberrant
individuals may be termed apparent mutants, since it may be con-
sidered uncertain whether they usually- arise by immediate mutation
or by segregation. The case acquires special significance because indi-
viduals belonging to the mutant types, although the species is known
to be typically Mendelian with respect to various characters, give
erratic heredity ratios suggestive of Oenothera.

At least eight types have been somewhat carefully studied, and six
of these have shown their heritability in progeny tests. Several other
types have been named, but for various reasons their distinctness is
more or less doubtful.

Some of the commoner types have each been produced by many
parents, and in several pure lines isolated from the original com-
mercial variety. The apparent mutants other than the early type com-
pose about four or five per cent of the progeny of Snowflake and early
parents, the separate types ranging down from about one per cent.

Most of the mutant types are in general inferior to Snowflake in
vigor, and the difference in development is greatly increased by certain
unfavorable environmental conditions. The proportion of apparent
mutants in cultures from Snowflake parents appears to be definitely
lower where germination is comparatively poor.

The mutant types differ from Snowflake and from each other in
various respects. The early type is practically a smaller and earlier
Snowflake. The other mutant types, on the other hand, differ markedly
from Snowflake in vigor, fertility, and various form and size char-
acters. Each type is named from some conspicuous characteristic
difference from Snowflake, but usually various other differences can
readily be found.

Somewhat extensive progeny tests have been made for five of the
mutant types, and a little evidence secured for two or three other types.

[1591

302 MISCELLANEOUS STUDIES

The early type is probably due to a single dominant mutant factor
segregating normally from the corresponding Snowflake factor; the
quantitative nature of its differences from Snowflake, however, makes
positive determination of this point a matter of great difficulty.

At least five other types plainly reproduce themselves, but about
50 to 70 per cent of the progeny are usually Snowflake; no true-
breeding individual of any generation of any of these types has yet
been tested. Genetic work with most of these types has been much
hampered or even prevented by low vigor and fecundity, and the
aggregate data from progeny of parents of four types strongly indi-
cate selective viability at germination. It has been determined by
crossing that in three of the types the mutant factor (or factors) is
carried both by eggs and by sperms. From these facts it seems prob-
able that homozygotes of the mutant types are non-viable, and that
severe selective elimination occurs during embryonic development;
or, in other words, that the mutant factor is imperfectly recessive for
a lethal effect.

In three types there appears to be linkage with the factor pair for
singleness and doubleness of flowers, the mutant factor being coupled
with singleness in the tested apparent mutants of two types, and with
doubleness in the third type. AA7ith two other types these factors
seem to be independent. No reversal of coupling has been found in
later generations of the former two types, but on the scheme presented
crossover singles should be scarce.

For one type (crenate-leaved) a hypothesis based on the facts stated
gives very closely the ratio obtained from selfed parents. Reciprocal
crosses with Snowflake conform less closely to the requirements of the
hypothesis, but do not definitely contradict it. The slender type,
which shows similar apparent linkage, seems to disagree definitely
with the hypothesis; there is strong evidence, however, that slender

individuals may differ genetically among themselves.

r

A more complex scheme providing also for the usual origin of these
types from Snowflake by segregation is briefly outlined.

The selfing ratios are very suggestive of duplication of a chromo-
some (non-disjunction), as in Oenothera lata, but it is hard to
reconcile the cases of apparent linkage with this hypothesis. It seems
probable that these three linked types have originated and are trans-
mitted in the same general way as the double-flowered type, and that
all of these four mutant factors (including double) represent changes
of some sort within a chromosome of the same .pair, which may be

[160]

MUTATION IN MATTHIOLA 303

numbered I. Miss Saunder's work shows that two flower-color factors
also belong to this linked group.

The large-leaved type strikingly resembles Oenothera gigas, and it
may prove to be triploid in nuclear constitution. In that case segrega-
tion may be irregular and genotypically intermediate individuals may
be more or less frequently produced.

It is probable that further study of these types will help to explain
the remarkable genetic behavior of Oenothera and of Citrus.

LITERATURE CITED*

ATKINSON, GEORGE F.

1917. Quadruple hybrids in the F1 generation from Oenothera nutans and

Oenotltera pycnocarpa, with F.. generations and back- and inter-
crosses. Genetics, vol. 2, pp. 213-260, 11 tables, 1 diagr., 15 figs.

BABCOCK, ERNEST B.

1918. The role of factor mutations in evolution. Am. Naturalist, vol. 52,

pp. 116-128.
BARTLETT, H. IT.

1917. Mutation in Matthiola annua, a "Mendelizing" species. [A review

of paper of same title by H. B. Frost.] Bot. Gaz., vol. 63, pp. 82-83.
BATESON, WLLIAM, AND SAUNDERS, EDITH R.

1902. Experimental studies in the physiology of heredity. I. Experiments
with plants. Matthiola. III. Discussion. Roy. Soc. London, Re-
ports to the Evolution Committee, vol. 1, pp. 32-87, 125-160, 15
tables.
BATESON WILLIAM, SAUNDERS, EDITH R., AND PUNNETT, REGINALD C.

1905. Experimental studies in the physiology of heredity. Matthiola. Roy.

Soc. London, Reports to the Evolution Committee, vol. 2, pp. 5-44,
tables.

1906. Ibid. Stocks. Ibid., vol. 3, pp. 38-53, 4 tables, 2 figs.

BATESON, WILLIAM, SAUNDERS, EDITH R., PUNNETT, REGINALD C., AND KILLBY

(Miss) H. B.

1908. Experimental studies in the physiology of heredity. Stocks. Roy.
Soc. London, Reports to the Evolution Committee, vol. 4, pp. 35-40,
3 tables.
BELLING, JOHN.

1914. The mode of inheritance of semi-sterility in the offspring of certain
hybrid plants. Zeitsehr. f. indukt. Abstam.- u. Vererbungsl., vol. 12,
pp. 303-342, tables, 17 figs.
BI.AKESLEE, ALBERT F., AND AVERY, B. T. JR.

1919. Mutations in the jimson weed. Jour. Heredity, vol. 10, pp. 111-120,

11 figs.

* An asterisk prefixed to the date indicates that the paper cited has not been
seen by the present writer.

[1611

304 MISCELLANEOUS 8TUDIK*

COBRENS, CARL E.

1900. fiber Levkojenbastarde. Zur Kenntniss der Grenzen der Mendel 'schen

Regeln. Bot. Centralbl., vol. 84, pp. 97-113.
1902. Scheinbare Ausnahme von der Mendel 'schen Spaltungsregel fur Bas-

tarde. Deutsch. bot. Ges., Ber., vol. 20, pp. 159-172, 4 tables.

DAVIS, BRADLEY, M.

1917. A criticism of the evidence for the mutation theory of de Vries from

the behavior of species of Oenothera in crosses and in selfed lines.

Nat. Acad. Sci., Proc., vol. 3, pp. 705-710.
FROST, HOWARD B.

1911. Variation as related to the temperature environment. Am. Breeders'

Assoc., Ann. Bept., vol. 6, pp. 384-395, 4 tables, 4 charts.

1912. The origin of an early variety of Matthiola by mutation. Ibid.,

vol. 8, pp. 536-545, 5 tables.

1915. The inheritance of doubleness in Matthiola and Petunia. I. The

hypotheses. Am. Naturalist, vol. 49, pp. 623-636, 1 fig., 2 diagr.

1916. Mutation in Matthiola annua, a "Mendelizing" species. Am. Jour.

Bot., vol. 7, pp. 377-383, 3 figs.

1917. A method of numbering plants in pedigree cultures. Am. Naturalist,

vol. 51, pp. 429-437.

GATES, E. EUGGLES.

1935. The mutation factor in evolution. London, Macmillan, xiv + 353 pp.,
1 map, 114 figs., bibl.

GOLDSCHMIDT, ElCHARD.

1913. Der Vererbungsmoclns der gefiillten Levkojenrassen als Fall geschlechts-

begrenzter Vererbung? Zeitschr. f. indukt. Abstain.- u. Verer-
bungsl., vol. 10, pp. 74-98, diagr.

1916. Nochmals iiber die Merogonie der Oenotherabastarde. Genetics, vol. 1,

pp. 348-353, 1 pi.

GOODSPEED, THOMAS H., AND CLAUSEN, E. E.

1917. Mendelian-factor differences versus reaction-system contrasts in hered-

ity. Am. Naturalist, vol. 51, pp. 31-46, 92-101.

HERIBERT-NILSSON, N.

*1915. Die Spaltungserscheinungen der Oenothera lamarclciana. Lunds Univ.
irsskrift, vol. 12, pp. 4-131. (Eeview by Ben C. Helmick in Bot.
Gaz., vol. 63, 1917, pp. 81-82.)

MULLER, HERMANN ,T.

1917. An Oenothcra-like case in Drosophila. Nat. Acad. Sci., Proc., vol. 3,

pp. 619-626.

1918. Genetic variability, twin hybrids and constant hybrids, in a case of

balanced lethal factors. Genetics, vol. 3, pp. 422-499', 1 table, 1 fig.,
1 diagr.

SAUNDERS, EDITH E.

1911. Further experiments on the inheritance of doubleness and other char-
acters in stocks. Jour. Genetics, vol. 1, pp. 303-376, 8 tables.
1911a. The breeding of double flowers. Fourth Intern. Conf. on Genetics,

Proc., pp. 397-405, diagr.
"1913. Double flowers. Eoy. Hort. Soc., Jour., vol. 38, pt. 3, pp. 469-482.

[16-21

MUTATION IN MATTHIOLA

305

1913a. On the mode of inheritance of certain characters in double-throwing
stocks. A reply. Zeitschr. f . indukt. Abstam.- u. Vererbungsl., vol.
10, pp. 297-310.

1915. A suggested explanation of the abnormally high records of doubles

quoted by growers of stocks (Matthiola). Jour. Genetics, vol. 5,
pp. 137-143, 3 tables.

1916. On selective partial sterility as an explanation of the behavior of the

double-throwing stock and the petunia. Am. Naturalist, vol. 50,
pp. 486-498.

SHULL, GEORGE H.

1914. Duplicate genes for capsule form in Bursa bursa-pastoris. Zeitschr.
f. indukt. Abstam.- u. Vererbungsl., vol. 12, pp. 97-149, 5 tables,

7 figs.

' ' STUDENT. ' '

1908. The probable error of a mean. Biometrika, vol. 6, pp. 1-25, tables,
4 diagr.

1917. Tables for estimating the probability that the mean of a unique series

of observations lies between — <*> and any given distance of the
mean of the population from which the sample is drawn. Ibid.,
. vol. 11, pp. 414-417, tables.

SWINGLE, WALTER T.

1911. Variation in first-generation hybrids (imperfect dominance): its pos-

sible explanation through zygotaxis. Fourth Intern. Conf. on
Genetics, Proc., pp. 381-393, 10 figs.

TSCHERMAK, ERICH VON.

*1904. Weitere Kreuzungsstudien an Erbsen, Levkojen u. Bohnen. Zeitschr.
f. d. landw. Versuchswesen in Oesterreich, 1904, pp. 533-638.

1912. Bastardierungsversuche an Erbsen, Levkojen, und Bohnen mit Riick-

sicht auf die Faktorenlehre. Zeitschr. f. indukt. Abstam.- u. Verer-
bungsl., vol. 7, pp. 81-234, tables.

WEBBER, HERBERT J.

1906. Pedigree records used in the plant-breeding work of the Department
of Agriculture, in L. H. Bailey, Plant Breeding CNew York, Mac-
millan), pp. 308-319.

DE VRIES, HUGO.

1906. Species and varieties: their origin by mutation. Ed. 2, Chicago, Open
Court Pub. Co., xviii + 847 pages.

1918. Twin hybrids of Oenothera hookeri T. and G. Genetics, vol. 3, pp. 397-

421, 14 tables.

1919. Oenothera rubrinervis, a half mutant. Bot. Gaz., vol. 67, pp. 1-26,

tables.

YULE, G. UDNY.

1911. An introduction to the theory of statistics. London, Charles Griffin &
Co., xiii + 376 pages, 53 figs.

fi6»l

306 MISCELLANEOUS STUDIES

PLATE 22

THE EARLY TYPE

Fig. 1. March 20, 1908. The single progeny of WG9. Plants from house M
to the reader's left from stake, from house W to right of stake, from house C
below. WG9-C10, the early apparent mutant, is the middle plant in the lower
row. The stake indicates inches.

Fig. 2. About May 1, 1908. WG9-C10 at the left, WG9-C9 (Snowflake) at
the right.

[164]

MUTATION IN MATTHIOLA

307

Fig. 1

Fig. 2

308 MISCELLANEOUS STUDIES

PLATE 23

THE EAKLY TYPE

Fig. 3. April 8, 1909. The single progeny of WG9-C9 (Snowflake); arrange-
ment as in figure 1.

Fig. 4. April 9, 1909. The single progeny of WG9-C10 (heterozygous early).
Warm-house plants partly at right of stake in lower row; arrangement other-
wise as in figure 3. Compare with figure 3, house by house.

[166]

MUTATION IN MATTHIOLA

309

Fig. 3

Fig. 4

| FROST ] PLATE 23

310 MISCELLANEOl'S STl'lHKS

PLATE 24

THE EARLY TYPE

Fig. 5. July 19, 1911. Lots 1 to 10, with lots 11 to 14 mostly in sight at
the right. Odd-numbered lot in nearer (west) half of each row.

Fig. 6. July 19, 1911. Lots 19 to 28, with lots 15 to 18 mostly in sight at
the left.

(168!

MUTATION IN MATTHIOLA

311

Fig. 5

Fig. 6

[ FROST 1 PLATE 24

312 MISCELLANEOUS STUDIES

PLATE 25

THE SMOOTH-LEAVED TYPE

Fig. 7. April 27, 1909. Smooth-leaved apparent mutants. Compare with
figures 3 and 4 as to earliness, noting the difference in date.

Fig. 8. May 29, 1914. Progeny of a smooth-leaved parent. Plant at right
Snowflake single, the others smooth.

[170]

MUTATION IN MAT TRIO LA

313

Fig. 7

Fig. 8

[ FROST ] PLATE 25

314 MISCELLANEOUS STfHlKS

PLATE 26

THE SMOOTH-LEAVED TYPE

Fig. 9. June 28, 1915. Progeny of a smooth-leaved parent. Smooth single
at left, Snowflake double at right.

Fig. 10. Same date and parent as with figure 9. From left to right: Snow-
flake double (also shown in figure 9), Snowflake single, smooth double.

1172]

MUTATION IN MATTHIOLA

315

Fig. 9

Fig. 10
[ FROST ] PLATE 26

316 MISCELLANEOUS STUDIES

PLATE 27

THE LARGE-LEAVED TYPE

Fig. 11. August 29, 1914. Progeny of a large-leaved parent (28a), near
the close of the hot Riverside summer. From left to right: large single, large
double, Snowflake single (two, the first injured by aphids).

11741

MUTATION IN MATTHIOLA

317

Fig. 11

f FROST 1 PLATE 27

3is M i. ST /•;/./.. i. \i-:nfs

PLATE 28

THE LARGE-LEAVED TYPE

Fig. 12. July 8, 1916. Progeny of a large-leaved parent. Middle plant
Snowflake; the rest large; all single.

Fig. 13. July 8, 1916. Progeny of a large-leaved parent, more than a month
older than those shown in figure 12. From left to right: large double, Snow-
flake double, large single.

[176]

MUTATION IN MATTHIOLA

319

Fig. 12

Fig. 13

FROST | PLATE 28

320 MISCELLANEOUS STUDIES

PLATE 29

THE CREXATE-LEAVED TYPE

Fig. 14. April 6, 1909. Crenate-leaved apparent mutants. Note the varia-
tion in leaf serration, and especially the slightness of the serration (or crenation)
with the one cool-house plant (below).

Fig. 15. April 14, 1911. Progeny of a crenate-leaved parent, grown in a
cool greenhouse. The first two plants at the right are Snowflake, the rest
crenate.

[178J

MUTATION IN MATTHIOLA

321

Fig. 14

Fig. 15
[ FROST ] PLATE 29

322 MISCELLANEOUS STWIKS

PLATE 30

THE CRENATE-I.EAVED TYPE

Fig. 16. July 8, 1916. Progeny of a crenate-leaved parent. From left to
right: crenate single (two), crenate double, Snowflake double.

Fig. 17. July 8, 1916. Snowflake X crenate-leaved, F,. From left to right:
smooth, Snowflake single, crenate double (two).

[1801

MUTATION IN MATTHIOLA

323

Fig. 16

Fig. 17
[ FROST 1 PLATE 30

324 MISCELLANEOUS STUDIES

PLATE 31

THE SLENDER TYPE

Fig. 18. April 27, 1909. Miscellaneous aberrant individuals, with two
typical Snowflake plants (third from the left above, second from the left
below). In upper row: second from left, narrow double; second from right,
slender double. In lower row at left, slender single (25b).

Fig. 19. April 14, 1911. Progeny of a slender parent (25b). Two at the
right Snowflake, the rest slender.

1182]

MUTATION IN MATTHIOLA

325

NT1R-W | " '

T T I

Fig. 19

[ FROST | PLATE 31

326 MISCELLANEOUS STI'DIK*

PLATE 32

THE SLENDER TYPE

Fig. 20. June 3, 1914. Progeny of slender parents. From left to right:
slender single, slender double, Snowflake double.

Fig. 21. July 7, 1916. Snowflake X slender, F,. Middle plant Snowflake;
the others slender; all single.

1184]

MUTATION IN MATTHIOLA

327

,. .

::i's MISCELLANEOUS STUDIES

PLATE 33

THE NARROW-LEAVED TYPE
Fig. 22. April 13, 1911. Narrow-leaved apparent mutants.

Fig. 23. June 3, 1914. A narrow-leaved apparent mutant among progeny
of a crenate-leaved parent. From left to right: narrow double, crenate single
(two).

[186]

329

Fig. 22

Fig. 23

FROST 1 PLATE 33

330 MISCELLANEOUS STUDIES

PLATE 34

THE NARROW-LEAVED AND SMALL-SMOOTH-LEAVED TYPES

Fig. 24. June 28, 1915. A narrow-leaved apparent mutant among F, progeny
from Snowflake X slender. Narrow double at left; the rest Snowflake single.

Fig. 25. April 14, 1911. Miscellaneous aberrant plants, some being apparent
mutants. From the left: first and fifth small-smooth, third stout dwarf, seventh
slender. See text.

[188]

X

MUTATION IN MATTHIOLA

331

Fig. 24

Fig. 25

[ FROST ] PLATE 34

332 MISCELLANEOUS STUDIES

PLATE 35

THE NARROW-DARK-LEAVED TYPE

Fig. 26. June 3, 1914. A narrow-dark-leaved apparent mutant among
progeny of a narrow-leaved parent. Third plant from left narrow-dark single;
the other three Snowflake double.

Fig. 27. June 28, 1915. Progeny of a "small-convex-leavedC?) " parent
(27a). From left to right: narrow-dark single, Snowflake double, smooth single.

[190]

MUTATION IN MATTHIOLA

333

Fig. 26

Fig. 27

[ FROST | PLATE 35

OCEAN TEMPERATURES

BY

GEORGE F. McEWEN

OCEAN TEMPERATURES, THEIR RELATION

TO SOLAR RADIATION AND

OCEANIC CIRCULATION

QUANTITATIVE COMPARISONS
OF CERTAIN EMPIRICAL RE-
SULTS WITH THOSE DEDUCED
BY PRINCIPLES AND METHODS
OF MATHEMATICAL PHYSICS

BY

GEORGE F. McEWEN

Oceanographer of the Scripps Institution for Biological Research
of the University of California

CONTENTS

PAGE

Introduction. The place of mathematical methods in researches on oceano-
graphic problems 337

Solar radiation and surface temperature, assuming the average rate of flow of
the water to be zero : 338

Preliminary discussion, and statement of certain generally accepted
conclusions as to the process by which the water gains and loses heat 338

Statement of assumptions, mathematical formulation of the problem and
its solution 339

Determination of the numerical values of the constants in the solution.... 344

Observed and theoretical lag of temperature maxima and minima behind
the radiation maxima and minima; comparison of computed and
observed normal temperatures 348

Numerical estimates of the coefficient of absorption of solar radiation
in sea water 350

Deduction of the change in surface temperature produced by a horizontal flow
of water 352

Preliminary discussion, statement of assumptions, and mathematical
formulation of the problem 352

Solution for the case in which the flow is constant 356

Solution for the case in which the flow is a periodic function of the time.... 359

Solution for the particular case in which the time interval is so small
that the solar radiation may be assumed to depend only upon the
' latitude 360

The rate of horizontal flow in the North Pacific off the California coast from
latitude 40° N to 30° N and in the North Atlantic off the west coast of Africa
from latitude 30° N to 20° N 362

The rate of flow off the coast of California deduced from surface tem-
peratures 362

The rate of flow deduced from temperature data compared with that ex-
pected from the empirically ascertained relation of winds to currents
and with direct observations on currents \ 363

The surface current prevailing for a short time interval near the north-
west coast of Africa, estimated from surface temperatures, com-
pared with direct observations and with results deduced from the
empirically ascertained relation of winds to currents 364

336 MISCELLANEOUS STUDIES

The relation of the temperature to time, depth and rate of vertical flow in the
depth interval from 40 to 600 meters 367

Statement of assumptions and mathematical formulation of the problem 367
Solution for the case in which the vertical flow is constant 369

Solution for the case in which the vertical flow is a periodic function of
the time 372

Numerical values of the constants in the solution, determined from tem-
perature observations in the Pacific near San Diego ... 374

Comparison of theoretical and observed monthly temperatures at depths
from 40 to 600 meters in the San Diego region 382

Solution of the problem of temperature reduction due to upwelling, with
numerical applications relative to the 40 meter level in the San Diego
region 382

Deduction of the change in surface temperatures due to a vertical flow of water
near the surface 388

Statement of assumptions and mathematical formulation of the problem
and solution for the case in which the flow is coastant 388

Solution for the case in which the flow is a periodic function of the time.... 391

Theoretical reduction of the surface temperature for each month in the San
Diego region due to upwelling, and comparison with observations 393

Deductions relative to oceanic circulation in the San Diego region, based on
Ekman's hydrodynamical theory 397

Deduction of the upwelling velocity in the San Diego region from the observed
relation of salinity to depth, and comparison with that deduced from tem-
perature data 406

Conclusion 415

Literature cited.... ....419-421

OCEAN TEMPEEATUEES 337

INTRODUCTION

THE PLACE OP MATHEMATICAL METHODS IN RESEARCHES ON
OCEANOGBAPHIC PROBLEMS

The present paper deals with the formulation and solution of
several quantitative problems suggested by data on ocean winds,
temperatures, and circulation. Before formulating these problems
a brief general discussion of the place of mathematical methods in
oceanographic researches is given.

The process of testing physical laws in the laboratory is greatly
facilitated by devising appropriate experiments in which the variables
are largely under the control of the investigator. Even under these
favorable conditions the actual phenomena are too complex for de-
tailed representation in a mathematical formula, and an appropriate
simplification by abstraction is required to formulate problems that
are amenable to mathematical treatment. This is true in a much
greater degree of the more complex phenomena occurring in nature;
yet a rigorous mathematical treatment of natural problems capable of
yielding results in agreement with observations, while in general more
difficult, is necessary and fully justifies the increasing attention being
given to terrestrial and cosmic physics.

The actual phenomena of heating and cooling of the water in the
ocean are far too complex to be considered in detail. Therefore, in
order to apply rigorous mathematical reasoning to these phenomena, it
is necessary to devise a comparatively simple ideal system which would
behave in essentially the same way as the actual one with reference to
the observations in question. Certain problems can then be formulated
definitely in such a way as to permit of the precise calculation of
results, the comparison of which with observations tests the practical
value of the abstract system.

It is fortunate for the problems considered in this paper that the
proper choice of the simple assumptions needed in devising the ideal
system is facilitated by certain general results of numerous and extended
ocean as well as laboratory observations. An abstract system founded
on such assumptions would in general agree much better with the
conditions in nature than one in which the assumptions were hypo-
thetical or carried over from some other field. Evidently deductions
from any group of simple assumptions cannot have the same degree

338 MISCELLANEOUS STUDIES

of certainty as direct observations, since the ideal system cannot con-
form accurately to all the details of the phenomena of the actual one.
However, if in this way a logical and reasonably accurate description
of a wide range of physical quantities is obtained there is good reason
to believe that deductions or predictions relative to quantities not yet
observed will be in agreement with the facts. This is especially true
if two or more lines of reasoning converge to the same conclusion.

When it is impossible or impracticable to make the appropriate
direct observations the theoretical results must be regarded as the
best estimates, even though it is not impossible that future observations
may show important deviations from theory. Finally, while the exist-
ing observations may be logically described by means of the ideal
system, and deductions based on it, extended results reached by apply-
ing purely deductive methods to the ideal system are not substitutes
for a correspondingly extended series of new observations. The neces-
sity for making observations will always exist.

SOLAR RADIATION AND SURFACE TEMPERATURE, ASSUMING THE

AVERAGE RATE OF FLOW OF THE WATER TO BE ZERO
Preliminary discussion, and statement of certain generally accepted

conclusions as to the way in ichich the water gains and loses heat.

In order to have a basis for estimating the effect of circulation on
ocean temperatures it is necessary to work out quantitatively the rate
at which the heat of the water is gained and lost under the more
simple condition of no flow. This will be done by devising an ideal
ocean, based on assumptions agreeing as nearly as possible with the
following conclusions which are founded on numerous and widely
extended ocean observations.

1. The primary source of heat is the radiant energy of the sun.
both direct and diffuse, that penetrates the water (Murray, 1912, p.
225, Gehrke, 1910, p. 67, Helland-Hansen, 1911-12, pp. 64-66).

2. Absorption of this radiation directly heats the water in the
upper layers (Nansen, 1913, pp. 21-22), and only a small fraction of
this radiant energy penetrates below 25 meters (Kriimmel, 1907. pp.
253-270, Helland-Hansen, 1911-12, pp. 65-68, and Knott. 1903-05).

3. There is always a complex vertical circulation (Helland-Hansen,
1911-12, p. 68, Gehrke, 1910, p. 68, Nansen, 1913, p. 21, Murray.
1912, p. 226) due to a lack of balance of the many forces acting on
the water particles. The resultant vertical flow through a finite sec-
tion due to this motion may be very small and may be either upward

OCEAN TEMPERATUBES 339

or downward. That is, at the same time, some portions of the water
are moving upward and others downward, thus tending to mix up the
water at different levels. In this problem the resultant of the upward
and downward flow will be assumed to be zero.

4. This "mixing process" is most intense in the layers nearest
the surface, owing to wave motion and other surface disturbances due
to wind, but is present in some degree at all depths (Gehrke, 1909,
p. 12, Murray, 1898, p. 127).

5. The amount of heat transferred from one level to another by
conduction through the water is a negligible fraction of that carried
by the water particles themselves as a result of the mixing process
(Gehrke, 1909, p. 12).

6. The mean annual rate of change of temperature with respect
to latitude is practically independent of the depth within the upper
hundred meters. This is revealed by a study of the average tempera-
tures of the North Pacific, tabulated with respect to latitude, longitude,
and depth (Schott, 1910, p. 14).

7. At the time of year when the surface temperature is a minimum
there is practically no variation of the temperature with respect to
depth in the upper thirty meters (McEwen, 1916, p. 272).

Statement of assumptions; mathematical formulation of the problem

and its solution.

Let Kl /! (L, t] equaZ Q^ the amount of radiant energy available
per month per unit area of horizontal surface at the latitude L and
time t, where Kl is proportional to the solar constant and fl (L,t)
is a function of the latitude L and the time t. Let Q equal the amount
of radiant energy used directly in heating the water, that is, the
amount passing into the water. Also let

y equal the distance in meters from the surface of the water, the

positive direction being downwards,
a equal the specific heat of sea water per unit volume,
& equal the temperature, centigrade.

t equal the time, the unit being 1 month, and t equal 1 for January,
01 equal a temperature assumed to depend only on the latitude and

depth y, and
/?! equal the average transmission coefficient of sea water for the

solar radiation, that is, the proportion of radiation at any level

that passes through unit thickness of water measured from that

level.

340 MISCELLANEOUS STUDIES

Since the solar radiation consists of a series of waves of varying
length, each having a different coefficient of transmission (Murray,
1912, p. 248, Kriimmel, 1907, p. 263), the use of a single average value
is only a simple approximation to the true relation.

No analysis will be attempted of the complex way in which the
heat in an element of volume, specified by given values of y and L,
is lost. This loss depends upon evaporation at the surface and on
the mixing process at all depths, and the rate of evaporation increases
as the temperature increases. Also heat tends to flow from regions

Fig. 1.

of high temperature to those of low temperature (Gehrke, 1910, p. 68).
It seems reasonable to suppose, therefore, that the rate of loss would be
greater, the greater the temperature.

Although the precise manner in which the rate of loss of heat
depends upon the temperature is not known, some definite form of
relation must be assumed in order to formulate th'e temperature
problem mathematically. For simplicity assume the rate of loss at
any depth to be proportional to (6 — 0J at that depth, where 0t is a
function of the depth and latitude only. Consider now the time rate
at which heat is gained and lost in a given rectangular element of
volume of unit cross section and thickness dy whose upper surface is
at the depth y (fig. 1).

The rate of change of heat in this volume element is evidently

t (1)

since the volume specific heat multiplied by the volume of the element
equals the change in the amount of heat per degree change of tem-
perature.

OCEAN TEMPERATURES 341

The rate of gain of heat in this element of volume due to the
absorption of solar radiation equals the difference between the rate
at which the radiant energy passes in through its upper surface and
out through its lower surface. At the upper surface the rate is
QPS and at the lower surface it is Qfl^*dy. Therefore the rate of
gain due to absorbed radiation is

(2)
since

= — (log ft

The rate of loss of heat will be assumed to be ka(0 — QJdy where
k is a function of y only.

Equating the rate of change of heat in the element to the rate of
gain from solar radiation less the rate of loss, we have the following
differential equation

(3)
which becomes

« _

ot
after division by ady.

Let L = Ll -f- x where Ll is a standard latitude chosen arbitrarily
and x is the distance in degrees from this position, x is positive for
latitudes higher than Ll and negative for lower latitudes. The function
f-i(L, t), (p. 339), then becomes f(x,t), which expresses the way in
which the radiation varies with respect to latitude and time. The
precise form of f(x,t] is unknown; however, estimates of the amount
of radiant energy available at the earth's surface made by Angot
(Hann, 1915, p. 40) can be closely approximated to within a ten-degree
interval of latitude by an expression of the form

Q1 = K1[ (ax -|- a2x) cos at -f- a3x -f- 1] (5)

7T

where a =— , and the coefficient of cos at is negative.
6

Assuming the amount of energy Q that enters the water to be pro-
portional to the amount available Ql

(6)

342 MISCELLANEOUS STUDIES

where the constant A' g A',, since the amount of energy used cannot
exceed the amount available. Equation (4) then becomes

.«.) (7)
Let

where B is a constant and

& = «-* (9)

where ft, is the absorption coefficient (Kriimmel, 1907, p. 263).
Equation (7) then reduces to the ordinary linear differential equation

- + kO=B[(al +o2x) cosaf + asZ + lJe-^ + fc^ (10)
Therefore

(11)

where F(x,y) is an arbitrary function of x and y. Integrating
equation (11) gives

. asm at -\-kcos at . Be~^v (l-fa,a-)
-a2x) - --- \- - --- \-0l

(12)
which can be readily transformed into

v — \suty yi**' «/ i^w I •» I

Va2-f fc2 k }

where

tan« = - (14)

k

and only the periodic part of the integral is retained.

If 0X is assumed to be independent of JT the latitude gradient g
of the mean annual temperature is, from equation (13),

. (15)

Therefore, since g is independent of y (sixth statement, p. 339)

k = kle~^ (16)

where fc, is a constant. That is,

g=2±

OCEAN TEMPERATURES 343

Corresponding to the time of year t0 when the surface temperature
has the minimum value 00 equation (13) becomes

PO = — 1- 6l — j- — T- \- - — (18)

Since the coefficient of cos (at — e) is negative (see p. 341) cos (oi0 — e)
must equal plus 1 ; therefore from the seventh general statement
(p. 339)

01=*t_:*^dbd^ (19)

Vo. + "

where 03 is a constant.

Making use of the results just found equation (13) becomes

[cOs(a*-e)-l]+^+r +*3 (20)

where

The small variation in temperature with respect to depth in the upper
six meters indicates that these upper water layers are very thoroughly
mixed (Michael and McEwen, 1915, 1916). Accordingly temperatures
in this 6 meter interval will be computed by using 3 meters, the
average value of the depth. That is, (y — 3) will be substituted for
y when the depth exceeds 6 meters and the constant value (6 — 3) =3
for all depths between 0 and 6 meters. From equation (20) it follows
that at the time of minimum temperature the temperature is inde-
pendent of the depth y in accordance with the seventh general state-
ment (p. 339). But the latitude gradient which is the part of the
coefficient of x in equation (20) not involving the time is

Bo,* e~^lV Ba,

. „ - _ _ I 7 '

This is not in accordance with the observed fact that the latitude
gradient is independent of y; but the relative error will depend upon
the ratio of the first term to the second term, and may not be important.
As will be seen later (p. 345) it proves to be a negligible error if we
add to $t of equation (19) the term Boxe-

Therefore equation (20) with this modification and the use of (y — 3)
for the depth y gives the approximate form of the relation between

344 MISCELLANEOUS STUDIES

temperature, time, depth, and latitude, in accordance with the assump-
tions and general results of temperature observations already stated.
The modified equation is

1--2 r , Baa B , Ba2xe-3t» (

6= *~ 2 -- [cos (atf— £)—!]+- — X + -T-+- +0-3

22 *2

(22)
Also from equations (19) and (20)

B , Ba.x , B(a1-\-a2x)e-^v-a)

cos (at~€} (23)

where

a a

tan c=-=-

k AV

Determination of the numerical values of the constants in the solution.

In applying mathematical methods to physical problems the
functional relation between the variables involves certain constants
which must be determined from observed values of these variables.
The constants in equation (22) are

ax a2 az a kl 03 B and b^
Also from equation (8) we have B = — — . The first four constants

cr

are found by fitting the function

•^i [ (ai + a2x) cos °-

(equation 5) to the estimated values of solar radiation, as given for
example by Angot's tables (Hann, 1915, p. 40). The next three
require observations on temperature. For example, they can be found
from the observed values of the normal annual range of temperature
at a series of latitudes and from the mean annual temperature at
these latitudes.

The coefficient of absorption ~b^ can be estimated from direct meas-
urements of the intensity of radiation in the ocean at different depths
(Grein, 1913), or from observations on water samples taken to the
laboratory (Petersen, 1912. p. 39). Also an indirect estimate can be
made from temperature and solar radiation data (pp. 350-352) by
means of equation (22). However, in the problem of surface tem-
perature in which the vertical flow is neglected the value of b^ is not
required.

OCEAN TEMPERATURES

345

Choosing 30° N for the standard latitude Zr1? the data on solar
radiation taken from Angot's tables and based on the atmospheric
transmission coefficient of 0.60 gives the following values of the first
four constants:

a=— .0128

= — .0159, a== 0.523.
b

Ol = — .47

A later and more accurate estimate of the radiant energy available at
the ocean's surface (Schmidt, 1915, p. 121) gives smaller values, on
the average, than those of Angot; also instead of the value — .0159

53

for az his results give — 01_ -m ~ — -0244, which will be used in

ZJ. i /\ JLU

in this paper.

Kriimmel (1907, p. 413) gives the observed normal mean annual
temperature and the normal annual temperature range for the surface
of the ocean corresponding to a series of latitudes. These values are
in part presented in the following table.

range for the surface of the ocean at latitudes 10° N to 50 N

North latitude

10°

20°

30°

40°

50°

Mean Annual temperature
Annual range

27.2
2 2

25.4
3 6

21.3

6 7

14.1
10 2

7.9
8 4

Half Annual range

1.1

1.8

3.35

5.1

4.2

x

-20°

-10°

00°

+10°

+20°

From the tabulated values it follows that the gradient of the half
range is .175 and the gradient of the mean annual temperature is .72
from latitudes 30° N to 40° N. Let mx equal e~abl where 3 is the mean
depth of the upper homogeneous layer. From these values and
equation (22) we have

Va2

= -3.35 =

_ I / *l!5l

2\ a

==_i15==Ba2mi ["i I/ fe*m* Y "I
~2\ » / .

(24)

(25)

and

(26)

346 MISCELLANEOUS STUDIES

in which the last members of equations (24) and (25) are approxi-

/ k m \
matolv correct if ( — - — - ) is small.

Substituting the numerical values of ar a2 az and a from page 345
we have

B= Ejmk* = 29'5k* <27>

From equations (24) and (27) neglecting -( — — -) we have the first

2 ^ a. '

approximation

29.5*^ X 1-910 X (—.47) = -3.35 (28)

from which

3 35

=

'' 29.5 X 1.910 X. 47
Therefore

1 + 1/2 (JW±Y = 1.0292

fc^i = 1.0292 X -1265 = .1302 (29)

and

29.5 X. 1302 _ 3.84

Similarly from equations (25) and (27) we have

29.5 kimi X 1.910 X -0128 = .175

from which

17n
if m — __ • _ — 2427

" 29.5 X 1.190 X .0128 "

Therefore

i2= 1.1075

1clml = 1.1075 X .2427 = .2688 (31)

and

„ 29.5 X .2688 7.927

B = - = - . ( oZ )

Wtj Wlt

The difference between the values of B and kl found by these two
methods of computation is due to the fact that the ratio between the

OCEAN TEMPERATURES 347

half range of radiation at the standard latitude to the gradient of
the half range of the radiation is not quite consistent with the ratio
of the half range of temperature at that latitude to its gradient.

It seemed best to take the average of the two values and to apply
corrections to a^ and a2, making them agree with the temperature
data; then the difference between the corrected values a\ and a'2 and
the original values will indicate the magnitude of this discrepancy
between temperature data and solar radiation data. Accordingly

B=:

x x Wj

and

1/2 (.1302 -f .2688) = .1995 = fc^ = .20. (34)

From equations (24) and (25) the new values a\ and a'2 of ax
and a2 are

and

- 3.35 V4+ -3.35 (g5)

O.9

Substituting these numerical valuations in equation (22) and deter-
mining 03 from the observed surface temperature 21?3 when x equals 0
we have, expressing the angle in degrees

M_ (—.318 — .0166*0 cr^y-v

[COS (30t — e)— 1]

V
.

274 + .04

' | 1T.c (37)

V-274 + .04
where

tan c=: 2.62 ,W* . (38)

g-bl(j/-3)

For the surface temperature, put y = 6 and we have

, = 5.9 (-.318 -.0166.) _6 _ 72i_5.9(.0166).

.DO .DO

(39)
+ 17.95 = — (3.35 + . 175oO cos (30^ — 69)° + 21.30 — .72a;.

348 MISCELLANEOUS STUDIES

In the same way the following values of the constants for the
latitude interval from 20° N to 30° N were determined :

a\ = — .343, o'2=— .0159, a3=— .0161, k.m^

—2.= the latitude gradient = — .41,

— .155 = the latitude gradient of the half range,

5.535 .523 m, m,

B=- -i and tan e-= -2^-^7^7 = 2.41

I ll-i _ I I {/

Substituting these numerical values in equation (22) gives for the
surface temperature

6 = — (3.35 + .155*) cos (30* — 67.5) ° + 21.30 — .41* (40)

Observed and theoretical lag of temperature maxima and minima

behind the radiation maxima and minima. Comparison

of computed and observed normal temperatures.

The value of e = 69°, corresponding to the latitude interval 30° N
to 40° N, was deduced from theory; and since 30° corresponds to

69

one month, 69° corresponds to —=2.3 months, the theoretical lag

oU

» ,, ( maximum ) . , ,. , ,, ( maximum ) j. ,•

of th-e j minimum f temperature behind the j minimum j radiation.

The theoretical time of the maximum temperature is therefore
8.3, or about halfway between August and September; while the
theoretical time of the minimum temperature is 2.3, or about halfway
between February and March. According to Kriimmel (1907, p. 407)
from numerous and extended oceanic observations the average time
of the lowest temperature is February (t = 2) and that of the highest
is August (t = 8).

Again, from the three curves (Makaroff, 1894, pi. 26) giving the
mean monthly surface temperature observed in the North Pacific
between latitudes 30° N and 45° N the minimum temperature occurred
when

* = 1.8, 3.0 and 2.5

and the maximum temperature occurred when
f = 8.1, 8.3 and 7.8

respectively. The average of the above values of t is 2.4 for the
minimum and 8.1 for the maximum. Thus the predicted value of the

349

lag agrees very closely with the observed value. Since this value was
computed from that of the period of the temperature change, which
is accurately known to be twelve months, and from the foregoing
determination of the value of k1ml this agreement between theory and
observation affords strong evidence in favor of the reliability of the
value .2 adopted for k^m^ which is an important constant in the
investigation of ocean currents presented later.

From numerous surface temperature observations in the Pacific at
long. 173° W, lat. 20° N (Puls, 1895, pis. 1-4), off Madeira in the
Atlantic, lat. 32° 30' N (Kriimmel, 1907, p. 407) and off Yokohama
and at long. 140° W in the Pacific, lat. 35° N (Kriimmel, 1907, p. 408;
Thorade, 1909, pis. 1-3) it was found that the mean annual tempera-
tures agreed well with the normal values for the latitude. Therefore
there is good reason to suppose that the condition giving rise to the
temperatures at these places approximates closely to the normal con-
dition. Thus a comparison between the theoretical monthly tempera-
tures given by equations (39) and (40) with the observed values will
give a still more detailed test of the theory (table 2).

TABLE 2
Theoretical and observed normal surface temperatures at a series of latitudes

Latitude 20° N

21° 18' N

30°N

32° 30' N

35° N

40°N

£
§
2

b

8
1

Observed

Difference

b

1

*Observed

Difference

b

o

O

H

b

o

I

H

Observed

Difference

b

o
I

J3
H

Observed

Difference

£>

1

1
2
3
4
5
6
7
8
9
10

24 °n

24 °1

+ °1

23 °3

21 °6

-1°7

18°7

16°6

18°0

+ 1°4

14°4

13°4

-1°0

10°1

23.6
24 0

24.2

24 7

+ .6

+ 7

22.9
23 3

20.8
21 6

-2.1
-1 7

18.0
18 2

15.7
16 0

17.2
17 1

+1.5
+ 1 1

13.6
13 8

13.8
13 4

+ .2
- .4

9.1
9.3

24.7
25 6

24.7
24 9

+ .0

- 7

24.1
25 ?

22.1
23 0

-2.0
-2 2

19.2
20 8

17.1

18 9

17.9

18 7

+0.8
-0 2

15.1
17 0

14.2

17 ?

- .9
+ 2

10.9
13.3

26 5

25 9

- 6

26 1

24 4

-1 7

2? 6

20 9

20 3

-0 6

19 2

18 5

- 7

15.9

26 8

26 3

- 5

26 5

25 3

-1 2

23 9

>•> \

21 8

-0 6

21 0

21 6

+ 6

18.1

27 2

26 7

- 5

26 9

25 8

-1 1

24 6

23 2

?:?: 9

-0 3

21 9

23 5

+1.6

19.1

26 8

26 7

- 1

26 5

25 1

-1.4

24 4

23 1

23 2

+0.1

21 6

2? 7

+ 1.1

18.9

26.1

26.4

+ .3

25.7

24.4

-1.3

23.3

21.9

22.1

+0.2

20.3

20.0

- .3

17.3

11

25.2

25.8

+ .6

24.6

23.0

-1.6

21.6

20.1

20.6

+0.5

18.4

17.2

-1.2

14.9

12

24.3

24.8

+ .5

23.7

21.1

-2.6

20.3

18.1

19.0

+0.9

16.2

16.4

+ .2

12.3

Mean

annual

25.4

25.5

+ .1

24.9

23.2

-1.7

21.3

19.5

19.9

+0.4

17.7

17.5

- .2

14.1

values

*Air temperatures at Honolulu (Monthly Weather Review, 1903, pp. 225-226).

350

MISCELLANEOUS STUDIES

It appeals that the theoretical results agree well with observation.
If the difference between the theoretical and observed mean annual
temperatures due to the average departure of the local conditions
from the normal is applied as a correction to the observed monthly
temperatures the agreement between theory and observation is very
close.

Numerical estimates of the coefficient of absorption of solar radiation

in sea water.

A lower limit of the value of the absorption coefficient bl can be
determined from quantities depending on surface temperatures and
the amount of solar radiation at the surface of the ocean by the fol-
lowing method. The value of Kl (p. 344) obtained from Angot's data
(Hann, 1915, p. 40) was

7f1 = 5.18X 107A

where A is the solar constant. The more accurate result of Schmidt's
later investigation (1915; p. 121) (p. 345) is

Since £ =

K 1== 217 X 30 X 10* = 3.255 X 107A.

Bml = 5.9 (p. 347)
.94X!0X5.95.54 6

g=

K

But since— is the ratio of the amount of energy supplied to the water

v
by solar radiation to the amount available at the surface, -=- < 1.

Therefore

K = 5.54 X 106 = 5.54

Kv ~m1bl X 3.255 X 107A ™ 32.55 wi^A <

Using the accepted value 2.00 for A we have
mjti = b^er^ > .0851.
TABLE 3

fc,

0

.05

.10

.12

.15

.20

.25

.30

be-**

0

.043

.074

.083

.096

.110

.118

.122

351

From the values of Z^e'361 given in table 3 it follows that &t > .12
or e~bl = & < .887 where {S^ equals the proportion of incident light
that passes through one meter of sea water. Direct observations of
the proportion of solar radiation passing through samples of sea water
taken from the Nordlichen Ostsee and the Bottensee (Petersen, 1912,
p. 39) give values of & varying from .60 to .86, which are less than the
upper limit .887 deduced from theory.

The variation with respect to depth of the heat absorbed by the
water tends to maintain a temperature gradient which would be
greater the smaller the transmission coefficient, and the mixing process
tends to reduce the gradient by transferring heat from warm to cooler
layers. That is, the rate at which heat is supplied to a given layer is
equal to that due to direct absorption of radiation plus the amount
due to the alternating vertical circulation of the water. But the rate
of gain of heat was assumed in the theory to be due entirely to the
absorption of radiation ; and therefore the estimate of the value of the
transmission coefficient deduced from observed temperatures at dif-
ferent depths would be larger than the true value. This conclusion is
confirmed by the following computation, based on temperature observa-
tions near San Diego (McEwen, 1916, pi. 26). The general equation
(22) (p. 344), is of the form

0 = R1 e-Mv-s) [cos (aj _€)_i] +R2

where Rt and R2 are constants, for a given latitude. Therefore,
Tj^g-Ms-a) eqliais the half range of temperature at the surface,
Tj^g-Mio-3) equais the half range at the depth of 10 meters and
[E1e~&i(6-3) — Rle-bi(10~3')] equals the difference between the mean
annual temperature at the surface and at the depth of 10 meters. If
there is a vertical flow (p. 374) the general temperature equation
reduces to the same form (equation 155, p. 390), and can therefore be
applied to temperatures in the San Diego region.

Substituting the observed average values of these quantities
(McEwen, 1916, pi. 26) gives

R^-31'! = 3.15 half range at surf ace (41)

Rl6-^ = 2.70 half range at depth of 10 meters (42)

Rl[e~a^1 — e~76i]=.40 difference in mean annual tem-
perature at surface and at 10 meters. (43)

352 MISCELLANEOUS STUDIES

From equations (41) and (42)

e«>i = |4?r =1-168 or 6, = .039

fi* I U

and

flje-36' — e-71*]=R1 (.1285) = A5 or #, = 3.50.

From equation (43) and the value, 3.5, already found for R1

e-** — er'r* = •£=• = .1142 and &1 = .034
o.D

which agrees approximately with the value .039 obtained from the
first equation. The average value .0365 should be used instead of the
large value of bl exceeding .12 (p. 351), in order to obtain the actual
rate at which the water gains heat as a result of both absorption and
mixture of water from other layers. That is, the rate of gain of
heat in the actual system takes place as if there were no such mixture
of the water and the coefficient of absorption were less than the true
value. Hence, as far as the rate of gain and loss of heat is concerned
we can substitute this more simple ideal system for the actual one.
We have now determined all of the constants of the original differential
equation (4), on page 341, which expresses the time rate of change of
heat in an element of volume, on the assumption that the average
flow, either vertically or horizontally, is zero. The modified tempera-
ture resulting from any additional factor, for example a current, can
be deduced by solving the above differential equation, to which has
been added the rate of change of heat due to this factor.

DEDUCTION OF THE CHANGE IN SURFACE TEMPERATURE PRODUCED
BY A HORIZONTAL FLOW OF WATER

Preliminary discussion. Statement of assumptions and mathematical
formulation of the problem.

It is well known, as stated by a prominent British hydrographer,
Wharton (1894, pp. 699-712), that "the most obvious phenomenon of
the ocean is the constant horizontal movement of its surface water,
which in many parts takes well defined directions. ' '

The work of both practical seamen and scientists has after many
years revealed the essential features of the main ocean currents, and
in a few limited regions a fairly detailed knowledge of the currents
has been obtained. However, all investigators agree that the esti-

OCEAN TEMPEEATUEES 353

mation of the direction and rate of flow of water in the ocean is
attended with many difficulties. Some of the methods of making such
estimates will now be briefly reviewed.

The most direct and widely used method is the comparison of the
position of a ship every noon determined from astronomical observa-
tion and from the log and course during the previous twenty-four
hours. The set of a current estimated in this way is subject to large
errors, unless special care is taken in making the observations. Under
ordinary conditions such estimates of currents less than ten miles in
twenty- four hours are quite uncertain (Kriimmel, 1911, p. 420).

Another method of studying currents is to use drift bottles enclos-
ing slips of paper on which to enter information as to when and
where they were found. A sufficient number of records of the initial
and final positions of these bottles and the corresponding time intervals
will, under favorable conditions, yield information especially as to
the average direction of the surface drift. This method is best
adapted to small enclosed seas as the bottles may then be easily
recovered soon after reaching the shores. For the open oceans it is
not satisfactory. Other floating objects, such as wrecks, icebergs, trees,
and plankton also furnish some information about the horizontal
circulation.

Under favorable conditions the current at a given place can be
measured directly by means of a current meter or by observing a
floating object designed to move with the current. In the open ocean
the difficulty of holding a ship in a reasonably fixed position usually
renders these methods impracticable.

Investigations of the causes of ocean currents and their relation
to these causes provide indirect methods of determining them. Ocean
currents are directly due to various external forces, the wind or
friction of a neighboring current, differences in pressure resulting
from evaporation, precipitation and differences in specific gravity, and
are modified by the deflecting force due to the earth's rotation and
by internal friction of the water. Thus, any theory of ocean currents
capable of yielding even a rough approximation to the quantitative
relations between the complex system of causes and the resulting
motion of the water would necessarily be highly complicated. As a
matter of fact, great difficulties always arise in attempts to establish
a connection between practical hydrography and theoretical hydro-
dynamics, and deductions of currents from their causes are quite
uncertain except in special cases in which the conditions in the ideal

354 MISCELLANEOUS STUDIES

problem agree well with those in nature. The application of theory
to practical problems is rendered especially difficult, first, because of
lack of knowledge of the frictional resistance to the motion of sea
water, and, second, because of the uncertainty regarding the current
produced by a wind of given velocity and direction.

One of the most important needs now is a comprehensive pro-
gramme of observations at sea, of the currents themselves and their
causes, supplemented by attempts to formulate hydrodynamical prob-
lems whose solution shall be consistent with the observations. .Much
credit is due to the pioneer investigators, Zoppritz, Mohn, Bjerknes,
Sandstrom, Ekman, Jacobsen, and others, for their development of
methods of dealing with such problems.

Another important aid to the determination of oceanic circulation
is found in the fact that a current consists of water particles tending
to preserve their temperature and salinity as they move along. Tln-s.-
characters change slowly and thus serve to depict the currents some-
what as do floating objects that are readily identified.

This part of the paper presents an attempt to develop, along the
line suggested in the following translation from Krummel (1911,
p. 439), a method of deducing currents from the temperature dis-
tribution :

No simple rule has been formulated for determining currents from tempera-
ture charts. But it is conceivable if not certain that a systematic investigation
of the so-called individual temperature changes will give a reliable basis for
the estimation of currents from temperatures. (We must distinguish between
the annual temperature range, corresponding to definite geographical positions,
and the practically uninvestigated temperature changes which one and the same
water particle undergoes along the great horizontal current systems. In a
continuous current, for example, the Gulf Stream, these individual temperature
changes which must be distinguished from changes at a given position may
run through the whole range from tropical heat to the freezing point.) The
problem is, however, very difficult, and a cursory comparison of the current
charts in the Atlantic as prepared from the Log Book of the "Seewarte"
reveals the great complexity of these closely inter-related phenomena. In
general, in connection with all water motions time is an all important factor.
Rapid and slow currents behave very differently as regards their heat content,
and can therefore give rise to widely different types of isotherms. Xo constant
angle between stream lines and isotherms can be proposed; the angle can vary
between 0° and 90°. The most frequent case is that of stream lines cutting
the isotherms obliquely.

In general, the rate of change of heat in an element of volume can
be expressed by adding to the right-hand member of the differential
equation (3) the rate of change due to other factors not considered on

OCEAN TEMPEBATUEES

355

page 339, and the solution of the new equation will give the tem-
perature under the new conditions. The rate of change of heat due
to a horizontal flow H of the water can be readily derived as follows :
consider a rectangular element of volume (fig. 2) of unit length per-
pendicular to the direction of flow and of breadth dz measured in the
direction of flow and thickness dy normal to the direction of flow.

Fig. 2.

Then the rate at which heat enters into the element less the rate at
which it is removed will be

HaOdy — Ha(0 + d6) dy = —HadBdij = —H<r^- dzdy (44)

oz

which is the time rate of change of heat in the element due to the flow
of water. Multiplying equation (3) by dz to make it apply to the
element of volume now considered and adding the above expression
for rate of change of heat gives the new equation

a—-dydz'= —
ot

— Ha-—dydz
oz

(45)

Dividing through by adydz and substituting the value of Q and
from page 341 we have

azx] cos at + a3x + 1] — fc(0 — 0J— H

which is the same as the temperature equation (10) with the term

356 MISCELLANEOUS STUDIES

(\n \
— H-— - ) added, x is the distance north or south from the latitude
bz '

chosen for reference, z is the distance from the same point measured
in the direction of flow, making an angle \f/ with the x direction ; there-
fore x equals nz where ^ is measured from the positive (north) direc-
tion of x, and n equals cos ^. Making this substitution in equation
(46) gives

l23l
oi oz

(47)
Solution for the case in which the flow is constant.

To solve equation (47) let

e = & -f e" + o"'

where 0' is the solution already found (equation 22, p. 344) correspond-
ing to H — Q, 0" a function of y and t only is to be determined, and
ff" is a general solution of the part left after substituting (ff + #")•
Substituting the value 6' + 6" + 6'" for 6 in equation (47) we have

A/J/ A/)" A/J'"

— +- - +TI— = B[(al + a2nz) cos at -f a3nz + 1] e~
ot ot ot

(48)

dz dz dz

From the definitions of 61, 0" and 0'" this equation reduces to

~dT" + dtf~ ~dz~ ~dz~

which can be broken up into two equations

and

_ i_ Tfff" I JJ —— Q (51 \

From equation (22) which gives the value of ff we have
off Ba,n . Ba.,nr

oz k'^/- ' - -- -•

sin at)] (52)

OCEAN TEMPERATURES 357

where g, K2, A^ and B: are constants having the following values :

A, = — =^= and B, = — =^=

\nl

Substituting the above value of- — in equation (50) gives

00

Art"

' — " ' — - [1— (^coirf-f^ana*)] [n

(53)

remembering that k equals k1e~bl(y'3) and for depths between zero and
six meters the constant value y equals 6 is to be used for ?/, while for
other depths the actual value of the depth is to be used for y (p. 343).
H, the horizontal velocity, may be any function of the time, but it
is assumed to be independent of z and y. Having in mind a numerical
application to be made later it will be convenient to let H equal the
periodic function of the time

H = H! ( 1 -f- «4 sin at -)- a5 cos at)

where H^ a4 and a5 are constants. Equation (53) then becomes the
ordinary linear differential equation of the first order in 6" and t

— #2r [1—

X (1 H- atsmat + ag cosaO (54)

Solving by the corresponding standard formula we have

X [1 + a4 sin at -f- a5 cos at] ekt + C \ dt (55)

where C is arbitrary but independent of t and z. Under these con-
ditions Ce~kt will evidently be included in a general solution of equa-
tion (51), and will therefore be neglected in the expression for 6".
Equation (55) can be directly integrated with the aid of well known
standard forms and the result for H equal to a constant velocity Hl is

H,n

]-

X [2ak sin a* + (A;2 — a2) cos at] (56)

358 MISCELLANEOUS STUDIES

Equation (51) in which .H\ ( 1 -f- a4 sin atf + OB cos a<) is substituted
for H becomes

t 4 5

Of OZ

In the special case where «4 = as = 0 the solution is

-f a4 sin at + a5 cos a*) T ~ =0

(58)

(2 2 \ / g 2 \
t — „ ° ) is an arbitrary function of it ^— ^ ) . This

solution can be easily verified by substitution in equation (57).

For a constant velocity H equals Hv, the general solution of
the differential equation (47) is the sum of the three quantities
V + 0" + &" • already found, and can be put in the form

B(al-\-a2nz}e-^-^ , , Basnz . B , Ba.nze-™ , , )

H t

k(z-zn) / 2 _ ~ \ 1

~ffT7 ^ -- j~^J \ (59)

Suppose the relation of the temperature to the time at a given
position 20 is known and that there is a constant horizontal velocity
Hl from that point in any given direction. From equation (59) the
temperature at any time and at any point along the stream line
down stream from the point z equals z0 can be found by giving the

arbitrary function / If— — ^-JL) such values that when z equals z0

6 = tf -f- 0" + ff" will equal the observed temperature, which is a
known function of the time at that point. All of the constants in
the equation are given on page 347. Therefore f(t') being known.
when t', the time at the position, z equals z0 is known, the arbitrary
function is determined. For a time t and a value (z — z0) of the
distance from z0 the expression

f(t') (60)

0-V)=(n (61)

since, in general, the function is determined by the values of the
independent variable

OCEAN TEMPERATURES 359

Solution for the case in which the flow is a periodic function

of the time.

If the flow is the periodic function of the time

(62)

the term 6' (equation 22, p. 344) will be the same as before, but 6" will
be given by equation (55) where a4 and a5 are retained. The integra-
tions can be readily performed with the aid of well known standard
forms and the result is

TT 11 ( / ~Rn p-6

JJ.-.II I / -DCt0t/

• )

s) 1

O 7 I *

2afc }. si
J

-2«X] sin2at

4 — 2a2a5 ] cos 2a^ ( 63 )

The solution of equation (57) when a4 and as are retained, found
by Lagrange's method, is

(64)
where f1 [ ] is an arbitrary function of

( t 4 COS at -| — 5Sin at } — 77

L\ a- a / #,J

This solution can be verified by substitution in equation (57), and
can be readily changed into the more suitable form

fc(z — Z0) fc(a« cos at — as sin oO r- / \

fl'" = e- #1 o " / ( f 4 COS a# + -5 Sin atf )

L V a a /

(65)

360 MISCELLANEOUS STUDIES

which reduces to equation (58) when a4 and a, equal zero. The
temperature at any time and place down stream from the position
where z equals z0 can be found if the relation of the temperature to
the time is known where z equals 20, by giving the arbitrary function
/[ ] values such that for z equals z0, e = 0'+6" + 6'" will equal
the observed temperature which is a known function of the time at
that position. Thus the arbitrary function is determined since its
value is known for a series of values of the independent variable

(f --- — cos at' -| -- - sin at' ) , using i' for the time where z equals Z0.
a a /

For any other value of the time V and for a distance (z — 20) down
stream

sin af-

where

f[(t— ^cos «* + -?*- si

L \ a a

=/[V — -^-cosaf + -^- sinaf 1 (66)

a a.

F/^ «4 , «-, A z — zn~]

\ (t- -cos at +-— sinafl -- =~

L \ a a //! — I

= (f— ^-cosaf + -^-sina*'). (67)

\ a a /

Solution for the particular case in which the time interval is so small
that the solar radiation may be assumed to depend

only on the latitude.

In certain cases it will be convenient to take a time interval so
short that the insolation may be regarded as independent of the time
and the current may be assumed to have a constant velocity from a
position z equals z0 where the temperature may be assumed constant.
Under these conditions the temperature at any point distant (z — z0)
down stream will be independent of the time if sufficient time has
elapsed for an element of volume of the water passing through the
position z0 and having the given constant temperature to move through
a distance equal to or greater than (z — z0). For this steady state,
equation (47) becomes

Ol}—Hi = 0 (68)

where 62 = 1 -j- aL cos at 1

&3 = a2 cos at: -}- a3

OCEAN TEMPERATURES 361

and t1 is the average of the values of t for the beginning and end of
the time interval. Let 6 = 6' -f- 6" where 6' is the solution when
Hl = 0. Then

>.«) (69)

i

Substituting (6'-\-6") inequation (68) gives

de" * d0'

(70)

For & use the normal value determined from the expression j

of equation (59) for t = t1 then for surface temperatures using the
value 6 for y (see page 343)

60'

where tan c =7-

k

This is consistent with equation (69) since 0X may be any function of z.
Equation (70) then becomes

Integrating equation (71) gives

_ fc(z— zn) a TJ
0" = Oe-^^-1 — -^ (72)

where 0 is arbitrary. Adding the two solutions ff and 6" gives

_ / ^ , , ^z0) Ba*mn , B
cos (a# — e) + - -- h' — i -- rr

(73)

where the expression j Ms the normal temperature. To deter-

mine the temperature at a given position, distant (z — z0) from the
initial position z0, give 0. such a value that the expression for 0 in
equation (73) will reduce to the given temperature when z = z0. Then
substitute this value and the given value of z in equation (73).

362 MISCELLANEOUS STUDIES

Consider two parallel stream lines, A and B, the velocity being
#A along the first and HB along the second, then the temperature in
A for any value of z minus the temperature in B for the same value
of z is

— Hz — Zo) _ fc(z Zn) a

—UB—+y-l(HB—HA)=M (74)

(75)

Denote HA—HB by AH, then

fc(z — z«) _ *r(z —

Also if -^— is small we have approximately

? _ 9 \ 1

-^

(76)

THE RATE OF HORIZONTAL FLOW IN THE NORTH PACIFIC OFF THE
CALIFORNIA COAST FROM LAT. 40° N TO 30° N AND IN THE NORTH
ATLANTIC OFF THE WEST COAST OF AFRICA FROM LAT. 30° N TO

20° N.

The rate of flow deduced from surface temperatures.

From the hydrographic charts (Thorade, 1909) of the region of
the Pacific off North America, it appears that that the average direc-
tion of the surface drift from Cape Mendocino, Lat. 40° N, does not
at any season differ greatly from a straight line determined by the
points, Lat. 40° N, Long. 124° W, and Lat. 30° N, Long. 126° W.
Assuming that there is a surface drift in this constant average direc-
tion which is proportional to the average wind velocity over this
course, will some numerical value of the drift account for the monthly
temperatures at the down-stream end of the line? From the
monthly isotherms worked out by Thorade (1909), the observed
mean monthly temperatures at any point of the region can be found.
From these observed monthly temperatures at the upper end of the
line and a mean value of the drift velocity determined by trial, the
temperatures at the down-stream end will be computed according to
the theory on page 359. A comparison of these theoretical tempera-
tures with the observed ones and of this theoretical value of the drift
with estimates made in other ways will indicate the practical value of
the theory. The observed temperatures taken from Thorade 's chart
(1909), and the numerical values of the other quantities computed

Month = i

Observed temp
Normal temper
0" (from equat

k

- (a^ cos at' — 05 sin

C' Ct

*t'— -4

Normal temper
tf + B"

(i — •— cosa# 4--

(* --- ^

^' 4. ^" _j_ ^'"
Observed temp
Computed min

* These two lin

OCEAN TEMPERATURES 363

from the theory (pp. 355-360) are condensed in table 4, in which
the following values of the constants are used: «4 = 0, a5 = — .6,
<*! = — .318, a3 = — .0244, a2 = — .0166, k —.20, and H = — 1.2
(1 — .6 cos 30° £) equals the drift, in degrees per month equals
1.0 to 3.8 miles in twenty-four hours. The mean wind velocity in miles
per hour over the course considered (Moore, 1908-11) is approximately
— .6 cos

The rate of flow deduced from temperature data compared with that
expected from the empirical relation of winds to currents and with
direct observations on currents.

As stated by Helland-Hansen in his paper on physical oceanography
(Murray and Hjort, 1912, p. 247) :

The wind may produce a current, particularly in the surface layers, thus
altering the direction and velocity of the existing current. We know very
little, however, about the relation between wind and current, through lack of
detailed observations, although the question is naturally of the first importance
from an oceanographical point of view, as well as from its bearings on the
conditions of everyday life. This is one of the principal tasks for the ocean-
ographer of the future; such observations are difficult to make, no doubt, but
with modern methods much can be done.

However, numerous observations of winds and currents have been
made. And, although the relation of wind and current varies with
the wind velocity, the latitude, coast line, depth, and distribution of
specific gravity, some progress has been made in estimating the drift
that a given wind velocity will produce. A careful investigation of
this question based on Ekman's theory (1905, 1906) and a large mass
of available data made by Thorade (1914) yielded the following
results. In case the coast is sufficiently distant and the effect of the
pressure gradient due to differences in specific gravity is small, the
drift will be directed at an angle of 45° to the right of the wind
direction in the northern hemisphere. The relation of the drift to the
wind velocity estimated by Thorade (1914, p. 387) is

, (77)

Vsin</> sec. hour

and

(78)

, ..

Vsin<£ sec. hour

where V is the wind velocity, H is the current in meters per second,
and <f> is the latitude.

364

MISCELLANEOUS STUDIES

The mean latitude of the drift computed from temperature data
(p. 362) is 35° and its direction was 45° to the right of the mean wind
velocity in accordance with Ekman's theory (1906) and Thorade's
estimate from observations (1914). From equation (78) and the
observed value of V (p. 363) the drift would be 3.9 miles in twenty-
four hours if it were due entirely to the observed winds, uninfluenced
by the coast and differences in specific gravity. This estimate is of the
same order as 2.4, that made from temperature observations (p. 363).
Again, direct observations of the drift having a southerly component
(Thorade, 1914, p. 283) near the head of the stream, Lat. 40° N to
50° N, gave the values presented in table 5.

TABLE 5
Observed surface drift, and values computed from temperature data

Month

4

5

10

Mean

Observed drift in 24 hours

3.69

2.05

2.41

2.6

No. of observations

21

24

54

Computed drift in 24 hours....

3.12

3.65

1.68

2.8

Thus the theoretical drift estimated from temperature data agrees
as well with the observations as could be expected. And it appears
from the comparisons made that estimates of the drift from tempera-
ture data will prove to be as reliable as those made by other methods.

The surface current during a short time interval near the northwest
coast of Africa, estimated from surface temperatures, and com-
pared with direct observations) and with results deduced from the
empirically ascertained relation of winds to currents.

From a series of direct measurements by means of a float designed
especially for the purpose (Schott ei al., 1914), the average flow
between latitudes 20° N and 28° N, off the west coast of Africa, was
found to be nearly parallel to the coast and toward the southwest.
These current measurements were accompanied by observations of
surface temperatures and winds, and the stations were distributed
along a line nearly parallel to the average surface drift and about
150 miles offshore. All of the observations were made during the

OCEAN TEMPERATURES 365

short time interval from June 2 to June 15, 1911, and are therefore
appropriate for the application of the theory developed on pages 360-
362 for estimating surface currents from temperatures.

The mean position during the three days, June 2, 3, and 12, was
Lat 30° N, Long. 14?6 W, and that during the three days, June 13,
14, and 15, was Lat. 24° N, Long. 17? 3 W. The distance between these
positions is 6.74, the unit being a degree of latitude, and the mean
surface temperatures were respectively 18? 32 and 19? 12, each value
being the average of eighteen observations. In equation (73), page
361, (z — 20) is the distance, measured in degrees of latitude in the
direction of the drift, and the initial position is in this case at Lat.
30° N. The direction of the drift was found to be to the south at an
angle of about 27° to the right (west) of the meridian, therefore the
change in latitude corresponding to the distance (z — 20) along the
line of the flow is (cos 27°) (z — £0) = .891 (z — 20). From these
values and the numerical values of the constants given on page 346,
equation (73) becomes

_ - .217(2 - Zn)

0 = 22.6— (.891) (.41) (z — z^+Be -- nl -- +(.891) (1.88JTJ

which gives the temperature at any point along the stream line, the
mean velocity being Hl degrees per month.

To determine Ht substitute the two mean temperatures with the
corresponding values of (z — ZQ), thus obtaining two equations

18.32 = - 22.6 +"0 + 1.685J3'1
and

. 1.461

19.12 = 22.6 + (.365) (6.74) + Oe~nT+ 1.685H,
Eliminating 6 gives the equation

1.461

-(4.28 + 1.6857IJ e~ST-\- 1.6S5H, = —5.94
from which the value of Hlt found by trial is

Hl = — 9.4 degrees per month = — .78 miles per hour.

Using this value of Hl the theoretical temperature at any point along
the stream line is

0 = 6.75 — .365(2 — *0) + 11.57e-°231<*~fo)

where (z — 20) is negative since the latitude decreases in the down-
stream direction.

366

MISCELLANEOUS STUDIES

Direct estimates of the drift were made at six stations along this
line from angular measurements relative to a float, the ship being
manoeuvered in such a way as to keep the sounding cable as nearly
vertical as possible. The values obtained at each station in the order
from north to south are 1.0, 0.7, 0.9, 0.8, 0.9, 1.3 miles per hour in a
southwesterly direction. Each value is the mean of about twenty-five

Fig. 3. Geometrical construction for determining the effect of a coast on
the surface current produced by wind.

observations. The components parallel to a line from the first to the
last station having the mean direction of the observed drift are 0.99,
0.7, 0.9, 0.7, 0.86, and 1.11 miles per hour, and the mean value is 0.88.

From the four estimates based on "dead reckoning" and the
position of the ship determined from astronomical observations at
noon, the drift appeared to be directed to the west of the direction
determined by the "float" method. The values are 0.4, 0.5, 0.4, 0.4
and the components parallel to the mean direction of the drift found
by the float method are 0.3, 0.2, 0.16, 0.19, the mean value is 0.21.

The wind blew steadily from the northeast, the observed velocities
in miles per hour being 28, 23, 28, 34, 3, 18, 13, 28, 28, 28, 34, and 34 ;
the mean is 25.

From equation (78), page 363, using 24°, the mean latitude of the
stream line for <f>, the drift due to a wind velocity of V miles per hour
would be .01975F miles per hour. Using the value 25 for V the un-

OCEAN TEMPEEATUSES

367

disturbed drift due to the wind would be 0.494 directed at an angle
of 45° to the right of the wind direction; this direction of drift is
nearly the same as that obtained from dead reckoning. If the same
wind velocity prevailed over the whole coastal belt, a correction to
the above estimate of the drift must be made (Ekman, 1906, p. 23).
The computation can be carried out graphically as follows (fig. 3).
Let OT be the direction of the wind, and OA represent in magnitude
and direction the "undisturbed drift" computed from equation (78).
If a circle is described through A tangent to OT, and a line AD is
drawn parallel to the coast then OD will represent in magnitude and
direction the corrected surface drift. In the case under consideration
the corrected estimate OD is twice the value of OA and makes an angle
of about 27° to the right (west) of the observed mean direction.
Therefore the component parallel to this observed direction is
(OD) cos 27° = 0.89(0Z>) = (0.89)2(0^4) = 1.78(0^4),
The results corresponding to various values of the wind (V) are
presented in the following list.

F

OA

OD

(OD) cos 27

13

.257

.51

.46

18

.356

.71

.63

25

.494

.99

.88

34

.672

1.34

1.20

Finally it is evident that the velocity of 0.78 miles per hour
deduced from surface temperatures agrees well with the estimates
made by the other methods.

THE RELATION OF TEMPERATURE TO TIME, DEPTH AND RATE OP
VERTICAL FLOW IN THE INTERVAL FROM 40 TO 600 METERS

Statement of assumptions and mathematical formulation of the

problem.

It has been found (p. 351) that the direct heating of the sea water by
the absorption of solar radiation is proportional to e~^v where bl > .12
and y is the depth in meters. Hence at the depth exceeding 40 meters
this direct heating effect would be less than 1 per cent of that at the
surface. Also the temperature range at that depth would bear the
same proportion to that at the surface if the variation in rate of
gain of heat were due only to the variation in this rate of absorption.

368 MISCELLANEOUS STUDIES

However, observation shows that there is a seasonal variation of 5°
at 40 meters and exceeding 1° at 100 meters (Murray, 1898, p. 127;
McEwen, 1916, p. 268) ; thus something other than the direct absorp-
tion of solar radiation must be the main factor in heating the water
of these lower levels.

These facts show that there must be a transfer of heat between
the upper and lower level, but the ordinary process of heat conduc-
tivity, as illustrated by laboratory experiments on still water, is
wholly inadequate to effect this transfer at a sufficiently rapid rate
(Wegemann, 1905<z, 1905&). It is now generally recognized that this
transfer of heat results from an alternating vertical (p. 338) circula-
tion of the water (Helland-Hansen, 1911-12, pp. 68, 69), in which at
any given instant certain portions of the water are moving upward
while others are moving downward. The resultant flow of a given
column of water may be either upward or downward, or may be zero.
Without analyzing the complicated process by which heat is trans-
ferred from one level to another in the ocean, it will be assumed to
be similar to ordinary conduction. But the coefficient of conductivity
corresponding to conditions in the ocean will depend mainly on the
intensity of the circulation or mixing process (Gehrke, 1910. p. 68;
Jacobsen, 1913. p. 71), and might be called the "coefficient of con-
vective conductivity" to distinguish it from the ordinary laboratory
coefficient. In the following investigation this coefficient of conduc-
tivity will be used and the direct effect of solar radiation will be
neglected. If the resultant vertical flow is zero, the well known partial
differential equation

="2 <79)

applies, where 6 is the temperature, t is the time, y is the distance
below the surface, and /t2 is the diffusivity, a constant proportional
to conductivity. If the resultant vertical flow is w it follows, as on
pages 354-355 that the time rate of change of temperature due to this

flow is ( — w -—] and the temperature equation then becomes

M 9 6-e 66 ,ftm

— = jr -T-T — w-7- (80)

dt dy dy

Equation (80) is a special case of the general equation of the conduc-
tivity in a moving medium ("Winkelmann, 1906, p. 444). Equation
(79), a special case of Fourtier's equation of the flow of heat in a

OCEAN TEMPEBATUEES 369

stationary medium, has been applied to the problem of temperature
distribution in the ocean by Wegemann (1905a, 1905&), using the
laboratory value of ft2, but the theoretical results were of an entirely
different order of magnitude from those given by observation.

Jacobsen (1913, p. 71) has successfully applied the equation of
the form (79) to some data on the distribution of salinity and cur-
rents in the sea near Denmark. He determined /*,2, the Mischungs-
intensitdt from field observations, using the idea that salt content,
quantity of motion, temperature, and other properties of sea water
vary because of the alternating changes in the position of the water
particles. The writer is, however, not aware of any application of
equation (80) to oceanographic problems.

Solution for the case in which the vertical flow is constant.

If the vertical velocity has the constant value wt then we have
(p. 368) to find a solution of the following linear partial differential
equation with constant coefficients

69 ,620 . 60

— — f^-—-\-w1-—=0 (81)

dt 6y2 a 6y

satisfying certain boundary conditions. To determine the temperature
at any depth, having given that at the upper level y = y2, we must
have a solution reducing to the given function of the time t (in this
paper it will be a periodic function of t) at the upper level and having
a given constant value at the lower boundary. A convenient method
of solution is to assume

e==MeaV+bt (82)

and substitute in equation (81). The result is

& -f WJL — fa2 = 0. (83)

Therefore eav+bt is a solution of equation (81) for all values of the
constant M and for all values of a and & satisfying equation (83) . Let

a = a1 ± &ti (84)

where ai and &x are real, then from equation (83) we have

b = [n*(a1* — l>l*) — wlal] ± [(2alAi2— wO&Ji (85)

370 MISCELLANEOUS STUDIES

and, if the solution is to be a periodic function of the time having

the period — where 04 is positive
«i

I^(al2 — blz) — wla1 = 0 (86)

(2a1^ — w1)b1 = a1 (87)

Solving equation (86) for al and equation (87) for bt we have

«/! ;+: v Wl " • ^7* "\ /QQN

«i= jri

and

7. - al

Since the temperature and the amplitude of the temperature
decrease as the depth increases, the exponent a^y and hence al must
be negative (y is positive in the direction from the upper surface
downward). Therefore only the negative sign is admissible before the
radical in equation (88) and

is definitely determined by given values of u\, ft2 and ±bl. Solving
equation (87) for aa gives

therefore because of equation (90) 01Lai must be negative, or &t

-

must be negative since ax is assumed to be positive. From equations
(90) and (91) we have

«1»=61X»+ VV (92)

and

where only the plus sign is admissible since &^ is necessarily positive.
Substituting this value of fe^ in equation (90) gives

a _ i
V

OCEAN IEMFESATVBE8 371

substituting this value of a^ in equation (89) gives

&i = !L-== (95)

2

which agrees with the result already found on page 370, that -^

PI
must be negative. From pages 369 and 370, 0—Meav+™, where

a — aizt&i* and b = ± a1= ± (2alfj.2 — wjb^

The solution of equation (81) is therefore

0 = j|fe«iv ± (fciiH-aie)i (96)

where M and at are arbitrary constants, a: and &! are given by equa-
tions (94) and (95) and the same value of ax is to be used with either
the plus or the minus sign before the expression in brackets. From
the properties of imaginary exponents equation (96) can be put in
the real periodic form

$=0*»{Almn(l>l<y + a1t)+Bieo* (fc^ + o^)} (97)

where A^ B1 and at are arbitrary constants.

Also since the differential equation is linear the sum of any number
of such expressions will be a solution. Therefore the following more
general expression

n = 00
0=

nv I An sin (bny + ant)+Bn cos (bny + ant) j (98)

is a solution, where An, Bn and an are arbitrary constants and an and
bn have the values

and

V2~an

Denoting - - \/-^f- by A,, and — l by A the following approximate
expression for an and &„ can be easily derived

a«=4+ (1 + *n2 + An4) A,, . (101)

6»=(1 — An2— 7«W4)A» (102)

372 MISCELLANEOUS STUDIES

W 2

where hn2= 5-5 — -z is small. If the velocity wt equals 0,

an=bn= ~\~n anc* eQuati°n (98) reduces to the well known

form

n = °o i . . — .
•«— \ — I JG*.\~ (. / 1 L \ /

4

(103)

A solution independent of the time results, if in equation (83) we

W

make & = 0 and a= — , or b = a = Q. The result is

e = Ce^ + D = C e>-y + D ;(104)

where C and D are arbitrary constants. This expression for 0 can be
added to the right-hand member of equation (98), giving the more
general solution

(105)

Solution for the case in which the vertical flow is a periodic
function of the time.

If the vertical velocity has the value w^=w1 [1 + rcos a£] the
general equation (80) becomes

— = 2— — w [I-1 J"66

Let

(108)

(109)

and substitute in equation (106), the result is

df(t] 2fl2 ^ _
dt

from which

a

where c is an arbitrary constant.

OCEAN TEMPEEATUEES 373

Let

a = a1±~b-Li or a? = a^ — &X2± 2a1&1i (HO)

then

— Oi^iT «W]t — ^sinat (HI)

the arbitrary constant of equation (109) being so chosen, that when
r equals zero, equation (111) reduces to equation (82) derived for a
constant velocity.

Rearranging the terms in the exponent of e in equation (111) gives

(112)
Let

I>1(2fjiza1 — w1)=a1 (113)

and

2/2 7i 2 \ /» * A / "1 "1 A \

as before (p. 370). Substituting in equation (112) and making use of
the properties of imaginary exponents gives

0 = M e °iy± ailj/t =*= sin at -sin at

(aj + 6^ — ^^sin a#) I

+ B cos [ajt + b,y — i^-sin at) Y (115)

where at and &, have the values given on page 370 and Alt Bl and al
are arbitrary. The values

6t = ax = 0 and al = 0 or A

are also consistent with equations (113) and (114) and lead to the
solution

D — c(l — e-~^r-

IT

where C and Z> are arbitrary constants and \ has the value —

'

374 MISCELLANEOUS STUDIES

The differential equation being linear, this solution can be added to
the right-hand member of equation (115) giving the solution

+ «--^.'n at | Al Sin fa,* + blV -^shl at~]
•4-BjCOB ajf-f-ftjl/ -- — — sinaf

D

fa^ + ftaj/ — ii^- si

cosa^ + ftaj/ — i- sin af (116)

which reduces to equation (105) corresponding to a constant velocity

(W T \

—sin at 1 is small

equation (116) can be transformed into the following approximate
form, retaining only the first powers of the small quantities.

2a

(cos (a — aj — btg) — cos (a + a^ + ^y) J — (B^

( sin (o + a^ + b,y) + sin (a — a^ — &,?/) J | (117)

Numerical values of the constants in the solution, determined from
temperature observations in the Pacific near San Diego.

It is well known that the waters of certain inshore regions, includ-
ing that off the west coast of North America, have a temperature
significantly below the normal for the latitude. Various explanations
of this phenomena off the California coast have been offered, but
(Holway, 1905, and McEwen, 1912, 1914, 1916) the only one so far
proposed that is consistent with all of the known facts is that of an
upward flow of cold water from lower levels.

OCEAN TEMPEKATUEES

375

Assuming this upward flow to be the only cause of the temperature
reduction, the theory developed on pages 372-374 will now be applied
to the series of temperature observations made off the Coronado Island
about twenty miles from San Diego (McEwen, 1916, pp. 267, 268 and
pi. 26). It follows from Ekman's theory (p. 401) that the vertical
velocity off the California coast is proportional to the component of
the wind velocity parallel to the coast, and decreases in magnitude as
the distance from the coast increases. For this reason the velocity
estimated from the temperature data mentioned above will be less
than that nearer the coast.

TABLE 6

Average of observed and computed monthly temperatures off San Diego
at a series of depths.

Month

1

2

3

4

5

6

7

' 8

9

10

11

19

mean

Depth in meters

o

15 0

14 3

14 6

15 9

16 1

17 8

IP 7

90 6

90 ?

18 8

16 8

15 5

17 0

0 C

o

15 0

14 6

14 5

14 7

15 4

16 6

18 9

IP 5

IP P

18 8

16 8

15 4

16 6

10 C

o

14 9

14 9

14 0

13 6

13 0

14 9

15 0

16 0

18 ?

IP 0

16 9

15 0

15 2

20 C

O

13 8

13 8

13 5

19 6

11 0

19 ?

13 0

14 0

17 8

18 0

15 8

14 5

14.2

30 C

O

13 4

13 4

13 0

11 0

10 5

10 8

11 8

12 7

14 8

16 8

15 4

14 0

13.1

40 C

o

13.4
13 1

11.9
13 0

10.6
19 3

9.9

10 5

9.9
10 0

10.6

10 5

11.9
11 0

13.4
1? 0

14.7
13 6

15.5
15 0

15.5
15 1

14.7

13 5

12 5

50 C

13 9

11 P

10 6

P 8

P 8

10 4

11 6

19 P

14 ?

14 P

15 0

14 4

o

1? 8

1? 6

11 7

10 0

P P

10 9

10 7

11 0

19 8

14 3

14 7

13 9

12 0

60 C
O
70 C

o

13.1
12.4
12.9
1? 0

11.8
12.1
11.7
11 7

10.6
11.0
10.6
10 7

9.9
9.9

9.8

p 8

9.7
9.8
9.6
P 7

10.2
10.0
10.0
P 8

11.2
10.5
10.9
10 3

12.5
10.8
12.0
10 7

13.7
12.2
13.2
11 8

14.4
13.8
13.9
13 0

14.6
14.1
14.2
13 7

14.1
12.9
13.8
19 7

11.6
11 3

80 C
O

12.7
11 7

11.6
11 3

10.6
10 3

9.8
P 6

9.5
P 6

9.8
P 8

10.6
10 ?,

11.6
10 5

12.7
11 ?,

13.5

1? 8

13.7

13 9

13.4
19 3

11.0

90 C

o

12.5
11 3

11.5
11 0

10.5
10 0

9.8
P 5

9.5
P 5

9.7

P 7

10.3
10 0

11.3
10 3

12.3
11 0

13.0
19 4

13.3
19 8

13.1
11 P

10 8

100 C

1? 9

11 4

10 4

P 7

P 4

P 5

10 1

10 P

11 P

19 6

13 0

19 8

0
150 C

o

10.3
11.2
P 5

9.9
10.7
P 0

9.2
10.0

8 8

9.0
9.4

8 7

9.0
9.0

8 8

9.2

8.9
8 P

9.5
9.1
P 1

9.8
9.6
P 4

10.2
10.2
P 7

10.9
10.9

10 9

11.0
11.3
10 4

10.7
11.4
10 0

9.9
9 4

200 C

o

10.1
8 4

9.9
8 1

9.5

8 3

9.1

8 9

8.7
8 9

8.5

8 9

8.5
8 4

8.7
8 7

9.1
P 0

9.6
P 4

10.0
P 4

10.2

8 7

8.6

300 C
O

8.5
7 8

8.5
7 7

8.4
7 6

8.2
7 5

8.0
7 6

7.8
7 6

7.7

7 7

7.7
7 8

7.8
8 0

8.0
8 4

80
.«

8 4

8.4

7 8

7.8

400 C

7 4

7 5

7 5

7 5

7 4

7 9

7 1

7 0

7 0

7 1

7 9

7 3

o

7.0

500 C

6 8

6 8

6 8

6 8

6 7

6 7

6 7

6 6

6 6

6 6

6 7

6 7

376

MISCELLANEOUS STUDIES

From table 6, which gives the observed temperature averages at
different depths and months, the constants of equation (116) or the
simpler approximate form equation (117) will now be determined.
The mean annual temperature 6m is given by the first two terms

}-D = em (iis)

(119)

which can be put in the linear form

e (em — Z>)= Log e€ + \y.

Assuming different values of D, plotting the results, and selecting the
value of D, for which the points fell most nearly on a straight line,
resulted in the following values of the constants:

= 5.6, 0 = 8.3, X = —.004

or

(120)

where y is the depth in meters.

The satisfactory agreement between the computed and observed
values of 0m, shown by table 7, proves that the form of the function
deduced from theory differs but little from the true form.

TABLE 7

Computed and observed mean annual temperatures at a series of depths from

40 to 700 meters

Depth

40

50

60

70

80

90

100

150

200

300

400

500

600

700

Om computed
@m observed

12?7
13 1

12?4
1? *i

12?1
1?! 0

11?9
11 6

11?6
11 3

11?4
11 0

11?2

10 8

10?2
Q 9

9?3
P 4

8?1
8 6

7?3

7 8

6? 7
7 0

6?4
6 3

6°0
5 5

Difference....

-.4

-.1

+ .1

+ .3

+ .3

+ .4

+ .4

+ .3

-.1

-.5

-.5

-.3

+ .1

+ .5

The time of minimum wind velocity is in December, that of the
maximum is in July (McEwen, 1912, p. 265), and the magnitude of
the wind velocity, and therefore the vertical velocity of the water,

is approximately proportional to 1 -|- r cos - where t = 12 corre-
sponds to the time halfway between June and July and r = 0.2.
Also since, as shown by table 6, the temperatures have the same period

as the wind, ai = a = ^ in equation (117). In order to determine
o

the remaining constants substract from the observed temperature for

OCEAN TEMPEEATVEES 377

each month and depth the observed mean annual temperature for
that depth. Then subtract the expression

-C

from each of these values using a provisional value of wt. Then fit

MeaiV cos (at -f- b^ — e')
the equivalent of the expression

e*>* ] Aj_ sin (at -f 6^)+ B1 cos (at -f b^y) 1

(p. 374) to these remainders, thus determining M, e, ax and &x.
The last part of equation (117) is neglected at first and its value

estimated later. Assuming w^ to be — 31 — r- the expression

- C ^£ (sin at} e^y becomes — .38 (sin -^ \ e-004" .

The following formulae corresponding to the special case when
a = ax = — are useful in determining a1? 61} Wj. and /u,2, and follow

W

from equations (89, 91. 94 and 95) remembering that \== — \

V?
.741 .524 .524

V"

(121)
.2619 .524

^T; rt = 2^ri

(123)

n

ax can readily be determined from the rate at which the annual
range varies with the depth. A was determined from the mean annual
temperature at different depths. These values substituted in equation
(122) give /*261 from which /** and bt can be found by substituting
in equations (123) and (122), and finally the product A^2 gives the
velocity wr*

* If X and 6, are given, the following equations derived from equations 90
and 95 can be used for computing the remaining quantities:

— \o,

378 MISCELLANEOUS STUDIES

The constants obtained in this way are listed below:
A = —.004, 0! = — .008, M2 = 7760, bt = — .00560, Wl = —

month

Ax=— -J~^=— .0058, M = 4AO, *'= —61°.
f" * 12

Assuming £ = 1 for January, and using the same origin for
determining the time in the expression for wind velocities, the expres-
sion for the temperature becomes

6 = 5.6 + 8.3e-004" + 4Ae-«08« cos (30* — .32t/ + 61) °
-,38e--004!'cos(30/4-75)0 (124)

remembering that C = 8.3 and D = 5.6 (p. 376).

The value of p2 = 7760, when expressed in c. g. s. units is

(100)2
7760 30X24X3600 =

which is about 25,000 times the laboratory value (Wegemann, 1905a,
p. 139). But this quantity /*2 is the same as the Mischungsintensitat
(Jacobsen, 1913, p. 71), which is a measure of the rate of transfer of
salts, heat, or other properties of sea water arising from the mixing
of water particles in the alternating circulation (p. 368). Suppose the
diffusion of salts and the molecular conductivity of heat to be negligible
in comparison to the rate of transfer due to the alternating or
reciprocal changes in the positions of the water particles: then as
Jacobsen (1913, p. 71) says, the value of this coefficient, the Mischungs-
intensitat, determined from any of the properties should be the same
under the same conditions. He found values ranging from 1.9 to 3.8
from observations on currents and the distribution of salinities in the
sea near Denmark. From observations on tidal currents and salinities
in a neighboring region he obtained the values ranging from 0.3 to
11.4. The value 30 obtained from temperatures in the San Diego
region is of the some order of magnitude, but the intensity of the
circulation in the two regions would probably be different, hence the
coefficients would be expected to differ.

The idea regarding the alternating motion of water in the ocean
held by Kriimmel and Ruppin (1905, p. 36) may be summarized as
follows : The coefficient of viscosity determined from laboratory experi-
ments, in which the motion of the water is slow and takes place along

OCEAN TEMPERATURES 379

parallel surfaces, varies from .008 to .02 for a wide range of tempera-
tures and salinities. But the idea of laminar flow no longer holds
when one considers the motion of whole volumes of water, hundreds
of meters in thickness, throughout which there is a pressure gradient,
as is often the case in oceanographic problems. In such cases one
should not use innere Reibung (viscosity), but Massenwiderstand
(hydraulic friction).

The character of the motion is no longer simply laminar (one of
simple 'sliding to and fro, parallel to a given plane) but the particles
of fluid leave their surfaces and move in vortices along stream lines
transverse to the laminar motion. Thus a much greater resistance is
developed. For example, the values of the coefficient of viscosity
obtained by Nansen in his ocean researches are 200 to 40,000 times
the laboratory value.

The same general idea has been successfully applied to certain
meteorological problems relative to wind, temperature, and humidity
by G. I. Taylor (1915). He found that the transfer, in a vertical
direction, of heat and water vapor in the atmosphere followed the law
of heat conductivity in solids, and that the effect of friction on the
motion of the air could be taken into account by substituting in the
general equations of motion a quantity called "eddy viscosity" for
the laboratory value of the coefficient of viscosity.

From the observed relation of air temperature to height off the
coast of Labrador, he obtained values of the coefficient; of ''eddy
conductivity" from .57 X 103 to 3.4 X 103, corresponding to wind
velocities varying from 2 to 3.4 Beaufort. Also from observations on
the relation of wind velocity and direction to height he obtained values
of the coefficient of eddy viscosity varying from .77 X 103 to 6.9 X 103.
The values of these coefficients are more than 10,000 times the labora-
tory values, the ratio being of the same order of magnitude as that
obtained for sea water.

In general, comparisons of his theoretical results with observations
indicated a very satisfactory agreement. An especially good agree-
ment was found between the predicted and the observed values of the
angle between the wind and the horizontal pressure gradient at
different levels.

380

MISCELLANEOUS STUDIES

JULY

350 Meters

I I

300 350 Meters

350 Meters

Figs. 4, 5, 6, 7, 8, and 9. Curves showing the theoretical relation of ocean
temperatures to depth in a region approximately eight miles west of the
Coronado Islands. The crosses ( + ) correspond to observed temperatures.

OCEAN TEMPERATURES

381

OCTOBER

50

100

150

2CO

fiO
300

100
350

150 200

400 Meters

250

300

350 Meters

NOVEMBER

150 200
400 Meters

300

350 Meters

DECEMBER

300

350 Meters

Figs. 10, 11, 12, 13, 14, and 15. Curves showing the theoretical relation of
ocean temperatures to depth in a region approximately eight miles west of the
Coronadp Islands. The crosses ( + ) correspond to observed temperatures.

382 MISCELLANEOUS STUDIES

Comparison of theoretical and observed monthly temperatures at

depths from 40 to 600 meters in the San Diego region.
The values computed from equation (124) and entered under the
observed temperatures given in table 6 are seen to agree well with the
mean of the observed values, thus proving the approximate correctness
of the form of the function deduced from theory. These computed
values and those from table 12 for the surface are also shown graph-
ically by figures 4 to 15, on which are entered a number of points
corresponding to actual observations (Michael and McEwen, 1915,
1916).

Solution of the problem of temperature reduction due to upwelling
with application relative to the 40 meter level in

the San Diego region.
In the relation of mean annual temperature to depth

D =--<!> (125)

deduced from the differential equation (80), A = -4 equals the

velocity divided by the diffusivity, but C and D are constants of
integration. From observations of the mean annual temperature at
a series of depths these constants can be determined as was done on
pages 372 to 376, and they correspond to the particular physical con-
ditions under which the observations were made. For the same value
of D, the deep water temperature, but a different value of one of the
physical conditions, say the velocity w^ what will the temperature <£
be? To answer this question it is necessary to know the relation of
each constant to the velocity w^. The relation of A to wl is known
and it remains to find the relation of C to u\.

In the limiting case in which A = 0, denote the new value of the
constants by C", D' and A'; then expanding the exponential gives

1' (126)

where B' = C'\r is the constant temperature gradient corresponding
to zero vertical velocit.

C=-/1(A) (127)

A

A) (128)

where A(0) =/2(0) = 1, since, as A = 0. C = C" and D= D'.

OCEAN TEMPERATURES 383

Substituting in equation (125) gives

>'/2(A) (129)

where the forms of the functions /t(A) and /2(A) are to be deter-
mined. At the greatest depth, yl for which the theory is valid, assume
the temperature to have the constant value ^ for all values of A.
What effect will a vertical velocity have on the temperature above this
level? The right-hand members of the equations (126) and (129) are
equal for y = y1, since <f>1 is assumed to be independent of A at that
depth, that is,

X)=B'y1 + D'+C' (130)

A

Therefore from equations (129) and (130)

^. (131)

Subtracting the general value of <j> given by equation (131) from
the particular value <f>' corresponding to the case of no upwelling
given by equation (126) gives

4>'_ + = B'(y — yj— f/1(A)(^_ex,l)=A</) (132)

A

the reduction in temperature due to the upwelling velocity wr= — /x2A.
It remains to determine B' and the form of the function /t (A).

The temperature change due to the variation of velocity with the
time was found to be approximately

(equation 116), where the velocity is

w=w1(l -\- r'cosat)

and the value of r in the remaining terms is neglected, for the fol-
lowing reasons. The values of the constants A^ and Bl depend mainly
on the seasonal variation in temperature due to radiation; if there
were no such variation they would be zero, in which case the variation
in temperature with respect to time would be due entirely to that of
the wind. The temperature reduction is therefore approximately

,_ t^lL 8in at

R + e^cl — e

384 MISCELLANEOUS STUDIES

where R is the constant average reduction. A variation of the velocity
from its minimum to its maximum value, that is, from 0 to 2u\ as at
varies from 0 to 2* produces a variation of temperature reduction from

to

l

that is, as the velocity varies from 0 to 2wl the temperature reduction
increases by the amount

n \,.f X«>)

= CeMe— — e

from which the temperature reduction due to the velocity wl is

(133)

The approximate temperature reduction deduced by two inde-
pendent methods is given by equations (132) and (133), respectively.
Equating these two values and using equation (127) gives

(134)

Solving for the unknown function /t(A) gives

A(y — yj

y-r
f ^il

(135)

Since A is assumed to be independent of y, the variation of the
right-hand member of equation (135) with respect to y is a measure
of the error in the two expressions for A<£. Also a comparison of the

AC

theoretical temperature gradient, £'=: ^--^-- (equation 126), which

/i(A)

would be expected in case of no upwelling with observations of deep
water temperature in such regions affords an additional test of the
theory. The following values of the constants of equations (125) and

OCEAN TEMPERATURES

385

(133) computed on pages 376 to 378, (7 = 8.3, D = 5.6, \ = — .004,
w^ = — 31, y± = 600, — - = .237 and sinh — - = .24 are used in

a a

table 8, which gives the relation of /\(A) (equation 135) to y.

TABLE 8
The variation of /^(X) wni/i respect to y.

y

^(y-yi)

e*v— e**+ex»smh ( — )

V a /

AW

30

2.28

1.009

2.26

40

2.24

0.965

2.32

50

2.20

0.924

2.38

60

2.16

0.885

2.44

70

2.12

0.846

2.51

80

2.08

0.809

2.57

90

2.04

0.744

2.63

100

2.00

0.740

2.70

200

1.60

0.466

3.59

300

1.20

0.282

4.26

400

0.80

0.160

5.00

500

0.40

0.076

5.26

600

0.00

0.022

0.00

700

800

Table 8 shows the variation of f1(\) with respect to depth to be
less than 20 per cent in the depth interval from 30 to 100 meters,
hence the two methods (equation 134) of estimating A</> are in good
agreement within this interval. The value of A (A.), corresponding to
30 meters, the smallest value of y for which the theory is valid, gives
the best estimate of A (A), and hence of 5',, since the variation of /t(A)
with respect to y is least for small values of y. Substituting the

numerical values gives

\c,

(136)

the mean annual theoretical temperature gradients that would be
expected at latitude 32° 30' if there were no up welling and the other
conditions remained the same as those prevailing when the observations
in the San Diego region were made. The observed temperature
gradient at the depth of 600 meters not near shore would be inde-
pendent of seasonal variations and would be but little affected by
horizontal currents, and is in the depth interval of the observations
from which the constants of the theoretical formula for mean annual
temperatures were computed.

386

MISCELLANEOUS STUDIES

TABLE 9
Vertical temperature gradients in degrees per meter at the depth of 600 meters.

Indian Ocean

South Atlantic Ocean

North Atlantic Ocean

Lat. 30° to 35 S

Lat. 30° S

Lat. 30° N

-.004

-.0135

-.0025

-.006

-.0160

-.0080

-.007

-.0160

-.0090

-.0085

-.0165

-.0110

-.0095

-.0210

-.0150

-.0115

-.0260

-.0155

-.0130

-.0160

-.0140

-.0175

North Pacific Ocean

Lat. 30° to 35° N

-.0075

-.0140

-.0215

-.008

-.0150

-.0215

-.0095

-.0150

-.0220

-.0095

-.0150

-.0235

-.0100

-.0155

-.0245

-.0100

-.0155

-.0250

-.0105

-.0160

-.0260

-.0120

-.0170

-.0290

-.0120

-.0195

-.0305

-.0125

-.0205

-.0315

-.0350

Table 9 shows the observed temperature gradient at the depth of
600 meters and at the approximate latitude 32° 30', corresponding to
widely different positions between latitudes 30° S to 35° S and 30° N
to 35° N, and their average is probably a good approximation to the
normal gradient at 600 meters. The average of the 22 observations
in the Indian and Atlantic oceans (Schott, 1902, pp. 158-160) is
— .0126 degrees per meter, and the average of the 31 observations in
the North Pacific (Makaroff, 1894, pp. 456^64) is —.0179. The
theoretical result — .0148 agrees well with these observations.

The reduction of the mean annual temperature at a given level y
for a given velocity w\ is proportional to CeXv (equation 133). that is,
the reduction is proportional to the difference between the tempera-
ture at the depth y and the constant D. Therefore the temperature
reduction corresponding to a given month is proportional to the differ-
ence between the temperature at that time and the same constant D.
That is,

A4" <t>t — D <j>t — D rio7s

»\ T\ — , T\ V -1-0' /

OCEAN TEMPERATURES

387

where <f>t is the temperature at the time t, A<^ is the corresponding
reduction, <j>m is the mean annual temperature, and A<£ is the reduction
of the mean annual temperature. Substituting the numerical values
for y equals 40 meters, from page 385, equation (137) reduces to

A*t = 13\15566(8-3) ('852) ('24)= (^^) 1-7 = -227^-1.27

(138)

Substituting the observed values of <j>t at the depth y equals 40 from
table 6 in equation (138) gives the values of A<^ entered in the
second line of table 10.

TABLE 10
The monthly temperature reduction at the depth of 40 meters near San Diego.

t

1

2

3

4

5

6

7

8

9

10

11

12

*+t

A<£r

1.75
-.10

1.75

- 27

1.68
— 37

1.23
- 37

1.11

- 27

1.18
- 10

1.41
- 10

1.61

.27

2.09
.37

2.55
.37

2.23

.27

1.91
.10

*4

1 65

1 48

1 31

0 86

0 84

1 08

1 51

1 88

?. 46

?, W,

?, 50

2.01

But the vertical velocity is u\ ( 1 -|- .2 cos ^ Y and because of its

\WiT

variation with respect to time a correction equal to ey^ C ~^~ sin at
(equation 116) must be added to these values. The correction for
this case is

.38e--16 cos (30* + 75)° = .324 cos (30* + 75)° = A</>r

(last term of equation 124, p. 378), and the values are entered in the
third line of table 10. Finally the fourth line gives &<f>t + A</>r = A0(
the total temperature reduction at the depth y = 40 meters.

Computing C from equations (136) and (135) and substituting
the result in equation (133) gives the expression

(?/ — 600) (—.0148) sinh

(w 2 \
4Qg2J

6 7760

e 7760 — e 7760 4- sinh

e 7750

..
\4062 /

for the mean annual reduction in temperature at the depth y due
to the mean annual velocity of up welling ivl. The values of this ex-
pression corresponding to a series of values of n\ (expressed in meters
per month are presented in table 11 for the depth y = 40 meters.

388

MISCELLANEOUS STUDIES

TABLE 11

Theoretical temperature reduction at the depth of 40 meters corresponding to a
series of values of the vertical velocity «;,.

wl

0

10

20

30

40

50

60

70

80

90

100

Temperature
reduction

0°0

0°38

0°95

1°67

2°48

3°34

4°18

5°00

5°80

6°48

7°07

From plates 24 and 36 (McEwen, 1916), and from an examination
of the temperature data (Michael and McEwen, 1915, 1916) it appears
that the temperature reduction at the 40 meter level, inshore averages
from 2° to 3° more than that ten miles offshore. Also surface tem-
peratures as much as 9° below the normal have at times been found
close inshore in July or August, when the normal temperature is
about 23°, while the reduction of the surface temperature ten miles
offshore does not at any time exceed 3?5 (table 12). If, corresponding
to a temperature reduction of 1?7 ten miles offshore, the inshore
reduction is 1?7 plus 2?5 equals 4?2, the corresponding velocity of
upwelling inshore would, from table 11, be about twice as great as
that ten miles offshore.

DEDUCTION OF THE CHANGE IN SURFACE TEMPERATURES DUE TO A

VERTICAL FLOW OF WATER NEAR THE SURFACE

Statement of assumptions and mathematical formulation of the

problem, and solution for the case in ivhich the

flow is constant.

The temperature reduction at the 40 meter level due to upwelling
applies also to the surface water, but owing to the upwelling in this
upper level the surface temperature will be still more reduced. The
time rate of change of temperature in the case of no resultant flow
given by the differential equation (10) on page 342 plus the term

— ti'0 — will be the modified rate due to the vertical flow w0 as on
°6y

page 354. Hence the new differential equation corresponding to a
vertical flow near the surface is

dT

where k = k^

at + azx) cos a* -f a3x + 1] —

— w

oy

(139)

OCEAN TEMPEEATURES

389

In case w0 is constant, assume for the form of the solution

6 = fM tan at + f2(y)eoSat + fs(y)x + ft(y)+6i (140)

where ^ is independent of t and x. Substitute this expression in
equation (139), thus obtaining the following equations:

(sin «0 «/,(*) - k.
(cos aO [ -a/, (y) + B

ay

= 0
(142)
=0 (143)

(144)

in which # is regarded as a constant.

In case of no vertical velocity the variation of (6 — Ot) with respect
to time is small compared to its mean value, which is independent of
the time (p. 344). Therefore, assuming this to be true for the present
case, a good approximation would result if the constant part were
multiplied by the correct factor fete"6ij/ and the variable part by an
average value k. Making this change in equations (141) to (144)
we have the much simpler ones

(145)
(146)

(147)

From equations (145) and (146)

(149)

where

= k — ^& and tan 60=:^-.

390 MISCELLANEOUS STUDIES

From equation (147)

/s(y)=^ (150)

and from equation (148)

--5- equals — b^^-^v which gives

(152)

where ffz is a constant of integration. Substituting these values in
equation (149) gives

(153)
As on page 343

and since the change in temperature due to upwelling near the surface

is due to the gradient—- , only terms involving y should be functions
oy

of w0. Moreover, as w0 increases indefinitely the temperature at the
lower boundary of this upper layer should approach the constant value

where 0 — 0' ^ —\ — f

Therefore, as on page 344

(155)

7*

which reduces to equation (22), (p. 344), when w0 = 0, except that 6.2
takes the place of 0, and the average value of k2 takes the place of
[&,e~6l<1'~8)]2. The latter quantities are small compared to a2 and
therefore the difference between their values makes but little difference
in the result. This difference or error comes from the approximation
made in solving the differential equation (139), (p. 388). 6.2 is found
by subtracting the temperature reduction due to upwelling at and
below the surface (3 meter level) from 03, therefore in case of no

OCEAN TEMPEBATUEES

391

upwelling 02 equals 03. A second approximation to the solution of

\n

equation (139) can be found by substituting for — its value from

dy

the first approximation. The solution is then reduced to

i

/a2 +V k

- _ .
i 7, ~r

(156)

where

tan e., = •

1 i ™J0V1 \n'

k '

«2+ (hz)2
If t<;0 = 0, equation (156) is identical with equation (22).

Solution for the case in which the flow is a periodic function

of the time.

If the vertical velocity equals

W0[l -f-r COS (at — ej]

the -time rate of change of temperature can be obtained from the dif-
ferential equation (139) in which the last term is multiplied by

The new differential equation is therefore

l
dy

(157)

Let & be the solution already found when r = 0, and let 8" be the
correction due to r, then 0 = 0'-}- 0", and the result of substituting
in equation (157) is

dO" ^ 60" . .60' . .60"

-^- + A:0"+ w0— + w0r cos (a^ — O — + uy cos (at — ej— = 0

(158)

392 MISCELLANEOUS STUDIES

Equation (155), using y instead of (y — 3), gives approximately
ff and

(159)

66'

Substituting this value of - in equation (158) and letting

dy

0" = ve-^v results in the ordinary differential equation
dv

i kv — tt'o&jV — [tt'0rcos (at — ej] ' f_ — [f>na (a£ — e2) — 1]

/ 2 I ]• 2

— rw^vcos (a* — cj (160)

where v is a function of £ only. Let

k — ?/'(,&! = fc, and Mz = — — ' 1 _ 2 —
then equation (160) can be put in the form

)S (at — ej— ^ [COS (2at — e3)-|-COSc4] [

\

-f- rwj)^ cos (at — €x) (161)

where e2 4~ ei — c

When the vertical velocity is directed upward -uf0 is negative.

therefore

k2=(k — u?061)> | MJ0&i|

and the last term of equation (161) can be neglected in the first
approximation, which is

v = e-™ | — r¥2 f e *»' [cos (a* — Cl ) —-cos ( 2at — e3 ) — -cos J r« + C I

77- f Fa cos f, -\- A\ sin e,~| . FA'., cose, — a sin €]~|
: -^ \ I $+?- - J sin «' + [ -^+7^ Jeos a*

r2acosC3+A:2smc3"|
~L 2(4a2+^) JS1

[/>•„ cose, — 2asin€.,~| cose4) /ico\

2(tf+t.') Jeos ~~2^ [

The arbitrary constant C is 0, since from physical considerations
the solution must be a periodic function of the time. Substituting

OCEAN TEMPEEATUSES

393

(y — 3) for y in the exponential ve~blV and adding the result to 0 from
equation (156) gives the following approximate value of the tem-
perature in case the velocity of up welling is w0[l-|-rcos (at — ej]

COS (at — £,)— 1

rw0Bb1(al-{- a2

a cos c, -4- fc, sin e, ~| .
2 "2 L \sinat

a H-/C2

[A:, cos tj — a sin Cj ~1
a2 I /g ^

"2acosc, -4- fe0 sin to"! .
" 2. si
T*a J

cos e.

sc, — 2asinc,

]

,
COS 2^ e"

(163)

where

tan Co =

H-"2-^-

+-

(164)

THEORETICAL REDUCTION OF THE SURFACE TEMPERATURE FOR EACH
MONTH IN THE SAN DIEGO REGION, DUE TO UPWELLING ;
AND COMPARISON WITH OBSERVATIONS

The theoretical relation of ocean temperatures to time and depth
developed in pages 368-381 was found to agree well with observations
from 40 to 700 meters in the San Diego region. The theory developed
in pages 388-393 is valid for only the upper ten meters ; but no satis-
factory theory for the intermediate interval from 40 to 10 meters or
to the surface has been worked out. Now since the temperature reduc-
tion at the surface depends upon the upwelling in all three intervals,
it is necessary to estimate the reduction in this intermediate interval
for which we have no theory. A method of making this estimate is
included in the following plan of computing the theoretical tempera-
ture reduction at the surface.

394 MISCELLANEOUS STUDIES

An explanation of symbols used in making the computations will
be given for reference:

ff equals the normal surface temperature.

6 equals the theoretical surface temperature when the effect of
upwelling is considered.

~& equals the mean annual normal temperature at the surface.

Ob equals the constant temperature at the depth 600.

A07 equals the total theoretical reduction of the surface temperature.

A0'a equals the theoretical reduction of the temperature at the
depth of 3 meters which corresponds to surface conditions
(p. 343).

A<£' equals the theoretical reduction of the surface temperature due
to upwelling in the interval from 3 to 600 meters..

A<£' equals the mean annual temperature reduction at the 3 meter
level due to upwelling in the interval from 3 to 600 meters.

A<£8 equals the mean annual temperature reduction due to up-
welling at the 3 meter level.

A0' equals the total theoretical reduction of the mean annual tem-
perature at the depth of 3 meters.

A^r'3 equals the mean annual temperature reduction due to up well-
ing in the interval from 40 to 3 meters.

Ai£'40 equals the mean annual temperature reduction due to up-
welling in the interval from 600 to 40 meters.

Throughout the interval from 100 to 40 meters the temperature
reduction increases at a constant rate by the amount .36 (table 8),
the velocity is practically constant (=w^), and the mean annual

2 3

temperature gradient is -^r. In the interval from 40 to 3 meters the

bO

1 7 Q 1 Q 1 CJ ()

mean annual gradient is - ' — = ^- (table 6), but the mean

o7 01

velocity is approximately = (1 -(- q)v\ where qu\ is the velocity nt

£t

the depth 3 meters. Assuming provisionally that g = 0.1 (p. 403) the
mean velocity in this interval is .55n\. The temperature reduction
in any depth interval is proportional to the length of the interval, as
was shown to be the case in the interval from 100 to 40 meters, and

OCEAN TEMPEBATUBES 395

is proportional to the temperature gradient (p. 386). Therefore if the
velocity were the same in the interval from 40 to 3 meters, as at the
levels below 40 meters, the following relation would hold

.36 ~/2.3\6()~2.3

_ /
\

_ _ _

~

/2.3\
\60 /

But using the provisional estimate 0.55^ of the mean velocity in this
interval we have Ai//3 = .55 X -61 = .34. Using the principle (p. 386)
that for a given velocity the temperature reductions are proportional
to the temperature gradients

(165)

Solving for A<£' gives

(/)/ /) A /)' \

" - ^6 — ^" s\ i A~7/\ i-\cc\

- — — =— ) (A<£ ) (166)

ff—e*—*ff. )

From page 387, and the value of Ai//3 we have A^'=1.7 + .34= 2.04.
From page 347, the normal temperature is

0'=— 3.79 cos (30# — 69) ° + 19.5 (167)

therefore

-3.79 cos (3Q(-69r 13.2 -A

since 0000=6?3 (p. 376).

The observed mean annual surface temperature is 17?0 (p. 375) but
the normal value less A^' equals (19?5 — 2?04) equals 17?46, which
is ?46 higher than the observed value. This indicates that there is a
still further temperature reduction of the surface temperature due
to upwelling at the 3 meter level. From page 390, 02 = 03 — A^>';
therefore, if A</>' is added to both members of equation (163) we can
replace 02 by 03, and the value of 0 -\- A<£' differs from the normal
surface temperature solely because of the upwelling at the 3 meter
level. Therefore 6' — (6 -f A<£') equals the temperature reduction
A0'8 due to upwelling at this depth. Since only the surface tempera-

396 MISCELLANEOUS STUDIES

tures are required k can be put equal to k, then k2 will equal k2.
From the value of ff (equation 24) and equation (163) we have for
the value of A0'8 = ff — (6 + A<£')

__ 1 _ ) _ rw0Bb^(a -f- aga: ) j cos c4 i"a cos <i + kz sin ci"[
" y^*,1) «'+V (~2fcT'~L a2+**2 J

, + ^2 sin «3~l
-' + *.') J

cos c, — a sin e,~\ , F2a cos e

sm " -

sin (2a0 + . : " cos (2c() r- (169)

Leaving w0 and therefore &2 and e, undetermined for the present"
equation (169) can be reduced to the form

.560 i
, cos(30* — c2) cos (30* — 69) | .0155w>(

i ««J-* i , _ _ . — • — - — — ccr\ i i

-560 ' (.274 + V)

( f.506 + .259A-,~| . , r.966fc, — .1361

| L .274 + VjSm at + [ .274 + V JC°S at

fl.046 cos €s -f k* sin c,"] . _ ^ , ffc, cos c3 — 1.046 sin e3"j
2(1.096 + ft>) " 2(1.096 + *,«) J

cos2af I (170)

using the following values of the remaining constants : a,^ = — .318,
a2= -.0166, Be-3^=5.9,Cl = 195, c = 69, e3=(€l+e2), c4=(e2— ej,
r=.2, a =.523, &, = .0365, (pp. 347, 351, 376), k = .2, k2 =

.2 — .0365w0, tan e2=T-(p. 394). Assume w0 equals — 3, then

K2

A^.= .448 — 3.78 cos (30* — 60) + 3.48 cos (30< — 69)

- .113 j 1.58 sin 30* + .441 cos 30* i =.448 —.69 cos 30* — .18 sin 30*

(171)

The values of the temperatures and their reductions are given in
table 12.

OCEAN TEMPEBATURES

397

TABLE 12
Normal surface temperatures and temperature reductions in the San Diego region.

t=

1

2

3

4

5

6

7

8

9

10

11

12

means

0'

16.55

15.75

15.95

17.11

18.91

20.86

22.45

23.25

23.05

21.89

20.09

18.14

19.50

A0S'

-.24

-.06

+ .27

+ .64

+ .96

+1.14

+1.14

+ .96

+ .63

+ .26

-.06

-.24

+ .45

A<£'

1.66

1.52

1.50

1.63

1.86

2.15

2.40

2.56

2.58

2.45

2.22

1.93

2.02

*A<£r

-.10

-.27

-.37

-.37

-.27

-.10

+ .10

+ .27

+ .37

+ .37

+ .27

+ .10

.00

•A0'

1.32

1.19

1.40

1.90

2.55

3.19

3.64

3.79

3.58

3.08

2.43

1.79

2.47

A0' obs.

1.55

1.35

1.35

1.91

2.81

3.06

2.75

2.65

2.85

3.09

3.29

2.64

2.50

0 comp.

15.23

14.56

14.55

15.21

16.36

17.67

18.81

19.46

19.47

18.81

17.66

16.35

17.03

0 obs.

15.00

14.30

14.60

15.20

16.10

17.80

19.70

20.60

20.20

18.80

16.80

15.50

17.00

Difference

+ .23

+ .26

-.05

+ .01

+ .26

-.13

-.89

-1.14

-.73

+ .01

+ .86

+ .85

+ .03

From line 3, table 10 (p. 387).

The agreement between the predicted monthly temperatures and
' the observed averages is very satisfactory. Moreover, from the rela-
tion of the vertical velocity to depth deduced from hydrodynamical
principles, and shown by table 13, the velocity w0 at the 3 meter level,
would be .116^, where w± is the velocity below the 40 meter level.
- That is, the value of w0 to be expected from that deduced from deep
water temperautres (p. 403) is .116 X ( — 31) equals — 3.6, which
agrees well with the value — 3.0 deduced from the surface temperature.

DEDUCTIONS KELATIVE TO OCEANIC CIRCULATION IN THE SAN DIEGO
•REGION BASED ON EKMAN 's HYDRODYNAMICAL THEORY

Zoppritz (1878) obtained some theoretical results relative to oceanic
circulation, from the general equations of motion of a viscous fluid,
and some of his conclusions have been widely used by oceanographers
and geographers ; but a critical examination made in the light of later
observations showed that his conclusions do not apply to conditions
found in nature. These erroneous conclusions are due to his failure
to take into account the deflecting force of the earth's rotation and
to his use of the laboratory value of the coefficient of viscosity. The
importance of the effect of the earth's rotation on currents in the air
and ocean was pointed out long ago by Hadley, Coriolis, and Ferrel ;
but with the exception of free currents, that is, currents moving by
their own inertia, the deflecting force due to the earth's rotation was
thought to be so small that it could be neglected until Bjerknes
(1901) first made clear the importance of this deflecting force in the
case of forced currents.

398 MISCELLANEOUS STUDIES

Ekman (1905) also used the general equations of the motion of a
viscous fluid, but included the deflecting force due to the earth's
rotation, and used in place of the coefficient of viscosity a constant
whose value was estimated by applying his formal solution of the
equations to field data. That is, he used a virtual value of the co-
efficient of viscosity in order to take into account the effect of the
irregular vortex motion which greatly increases the magnitude of the
mutual reaction between the adjacent water layers.

On the simple assumption that the depth of the region considered
is large and the coast is at a sufficiently great distance Ekman (1905,
p. 7) deduced the effect of a wind, constant in magnitude and direc-
tion, over the whole region. His results for the northern hemisphere
are, for the velocity of the water, perpendicular and parallel respec-
tively to that of the wind,

U' = V0e-™ cos (~ ay\ (172)

V = VQe-°y sin fc— ay\ (173)

Where V0 is the absolute velocity of the water at the surface y is the
depth below the surface and a is a constant. The value of a (Ekman.

1905, p. 6) is n = v* sm <fr where w is the angular velocity

"

of the earth, <f> is the latitude, p.2 is the virtual coefficient of viscosity,
and q is the density of the water. From equations (172) and (173)
it follows that the surface water velocity (where y equals zero) makes
an angle of 45 degrees to the right of the wind velocity, and the angle
increases as y increases. "When y has such a value (denoted by D)
that the water velocity has the opposite direction to that of the sur-
face, that is, when

f

(174)

the magnitude of the velocity is e~v or .043 times its surface value.
D is called the "depth of frictional influence," since the water velocity
below that depth produced by a wind over the open ocean is but a
small fractoin of that at the surface. From an estimate of the relation
between the wind velocity and its tangential pressure and the corre-

OCEAN TEMPEBATUBES

399

spending ocean current produced, Ekman (1905, p. 42) concluded that

.0127

the surface velocity V0 would be approximately
D would have the approximate value

meters

h and that

(175)

where h is the wind velocity in meters per second. Thus for a wind
velocity of ten miles per hour (5.1 meters per second) and at the

Fig. 16. Components of the water velocity U' and V perpendicular and
parallel respectively to the wind, and the velocity of Vc in a direction perpen-
dicular to the coast.

76X51
latitude 35° D would equal ' = 51 meters, and for a velocity

V sin <f>
of fifteen miles per hour D would equal 75 meters. Solving equation

(174) for fj.2 gives

2 = — in — •°000729 • )D2

IT* 7T*

which equals 217 in c. g. s. units, when D equals 7500 cm. and
<f> equals 35°.

The relation of vertical velocity to depth will now be deduced from
the results of Ekman 's theory.

Let V make an angle A with the coast (fig. 16) and U' the angle
(90-)- A) ° with the coast, then the velocity perpendicular to the
coast is, from equations (172) and (173),

F'sin A + Z/'sin (90 -f A) = Vc

400 MISCELLANEOUS STUDIES

where a velocity directed away from the coast is regarded as positive.
That is,

Yc="P0e~OI'jcosAcosQ —ay J-f sin Ash:Q -ay\ (

= V0e-a»cos(^ -ay— A) (176)

Let z equal the distance perpendicular to the coast, then Vc from
equation (176) is the limiting value of the velocity perpendicular to
the coast as z increases. Also, adjacent to the coast where z equals
zero, the velocity perpendicular to the coast must be zero. Let the
velocity perpendicular to the coast at the distance z be given by
the equation

V=Vcf(z) (177)

where /(0)=0 and f(z) =1 as z increases. The removal of
water at and near the surface due to a flow away from the coast
decreases the pressure at the lower levels, and gives rise to a com-
pensating flow of deep water toward the coast. Therefore in a coastal
region where the surface water flows away from the coast there is a
compensating upward flow of deep water.

Denoting the vertical velocity by W, the equation of continuity is

dz dy

neglecting the variation of the component parallel to the coast. There-
fore from equation (177), assuming

W = fl(z)f2(y) (179)

we have

/i(2)W=0 (180)

To solve equation (180) let

df(z}

(181)

dz
where M is a constant. Then from equation (176) and (180)

dy \4 /

OCEAN TEMPEEATUEES 401

from which we have

f2(y)= —MV0 I e-w cos ( ~ ay — A J dy + Cx

0 j sin A sin a?/ — cos A cos ay [ e-w -f Ox
Za ( )

i/y A/o"

' 2°a e"ayc°S (oy + A) +(71 (183)

where C^ is the constant of integration which must have the value

MV V~2~

- cos A in order that /2(0) may equal zero. From equa-

£i(t

tions (176, 177, 179, and 181) it follows that the horizontal and vertical
components of the velocity are respectively

V = f(z)V0e-a»cos( ^ -ay — A)

W = — fM \ — — =+ [e-°»coa(-—ay — \}dy \MV0
( V2« J \4 / )

(184)

df(z} ( f /* . \ , cos A )

~ay cos ( - -^- ay — A ) dy = V V0

e~ay cos (ay -f- A) — cos A f-

^ u,z

Since the horizontal velocity of the surface water is proportional to the
wind velocity (pp. 363, 398) it follows from equation (185) that the
vertical velocity of the water is also proportional to the wind velocity.

The differential equation of a stream line is in general -^-equals

(JLZ

w

the slope of the curve equals — equals

( f I" * Y* cosA )d/(z)

— •{ / e~ay cos ( ay — A }dy V —5 —

dy ( J \4 / V2a ) dz

._„(_._ (*_ _ay_A[f(2) (186)

which can be reduced to the exact differential equation

df(z)
/(*) "

whose solution is

e~av cos f ^ — ay — A J dy

j fe-»vcos(-— ay — \\dy- ^4=1 (187)
[cos A — e-** cos (ay -{- \)]f(z)= C2 (188)

402 MISCELLANEOUS STUDIES

where C2 is a constant of integration corresponding to a given stream
line. From equation (185) the upward flux through a horizontal area
of unit width and length z measured perpendicularly to the coast is

rWdz=—^

Jo

[cos A — e"aycos (az -f -, u \~, i ngo)

Rembering that /(O) =0 and /(«)=! as z =

\

I/O

0[cosA — e-^cos (ay + A)]/(«)

•jy^2__ L ^^_vv> (190)

and that the maximum numerical value for a given value of y is

I

Wdz = -V« [C°S A ~ e'"V e™ (191)

which approaches the value - ( ^) as V increases, the ratio R

of the upward flux through a horizontal area at the depth y of unit
width and length z measured perpendicularly to the coast to the total
upward flux is

I

J

Wdz

[cosA —

V cos A cos A

Therefore the parameter C2 in the equation of the stream line (equa-
tion 188) equals the ratio R multiplied by cos A, and the flux between
any two adjacent stream lines of a series in which the increments of
C2 are equal is constant.

The mean wrind velocity of the 5 degrees square of the U. S. Coast
Pilot Charts (Moore, 1908-11) west of San Diego was found to be
about fifteen miles per hour in a southeasterly direction, approxi-
mately parallel to the coast. Therefore for the San Diego region the
angle A in equation (185) is zero, and the vertical velocity at the
depth y is, from equation (185), proportional to [1 — e-"vcosay]

where a= ^ = ^ (p. 398). The values of this function are tabu-
D 75

lated with respect to depth in table 13.

OCEAN TEMPEBATUEES

403

TABLE 13
Tabulation of the function

I ^ _

\

- 75 CQS

75

y

Z?L Try

l-e-75 cos=|

I

y

I*. 7TW
l_g-75 COS=f-

75

0

0

18

.658

i

.043

19

.685

2

.085

20

.710

3

.116

21

.737

4

.166

22

.760

5

.207

23

.783

6

.246

24

.805

7

.288

25

.825

8

.325

30

.913

9

.361

35

.977

10

.399

40

1.02

11

.436

50

1.067

12

.471

60

1.065

13

.515

70

1.052

14

.535

80

1.034

15

.568

90

1.019

16

.600

100

1.007

17

.630

infinity

1.000

The relation of the velocity to depth was deduced from hydro-
dynamical considerations, but its relation to distance from the coast,
which requires the determination of the function f(z) (equation 177,
p. 400), did not result from the foregoing reasoning, but will now be
considered. From equation (177), page 400 /(0)=:0 and f(z)= 1
as z increases, and for large values of y off San Diego, equation (185)
becomes

T7 ,7JV~\

(193)

ay 2 dz

where (pp. 377, 388) W1 = — 31 meters per month where z equals 10
miles, and equals double that value or 62 meters per month, where
z equals zero. From page 363, for an average wind velocity of 15 miles
per hour or 7.5 meters per second

.0126 X 7.5
Vsin 35°

= .125 meters per second

404 MISCELLANEOUS STUDIES

equals 324,000 meters per month and a equals — (p. 402). Substi-

75

tuting these numerical values in equation (193) and expressing z in
miles gives

df(z) =.Q2W for 3 = 0,

dz
and

df(z)

= .0105 for 0 = 10. (194)

While these conditions, which f(z) and , must satisfy, do not

dz

determine the functions precisely, they suffice for a rough estimate.
The following form

f(z) = 1 _ kj-te — (1 — kje-*** (195)

has the value zero when z equals zero and approaches 1 as z increases
indefinitely for all positive values are of /^ and h2, and differentiating
with respect to z

(196)

The above expression for - satisfies the conditions expressed by

equation (194) for the following values of the constants found by trial :

7^ = 01, ^=.93, 7i2=.17, (1 — A:J=.07.

Therefore from equation (193) the vertical velocity at any depth ?/
equals

W = —2960 $1^1 = —296,0 (.0093e-01* — .0119e-17*) (197)
az

and from equation (192) the ratio of the upward flux within a dis-
tance z from the coast to the total flux is proportional to

/(z) = l_.93e-01* — .07e-1T* (198)

where z is the distance from the coast in miles. The values of /(«)

and -^5 — are tabulated with respect to z in table 14.
dz

OCEAN TEMPERATURES

405

TABLE 14
Tabulation of the functions f(z) and

dz

z

/(*)

dm

z

/(*)

df(z)

dz

dz

0

0

.0212

40

.380

.0062

1

.020

.0192

50

.430

.0057

2

.040

.0175

70

.530

.0046

3

.060

.0161

100

.660

.0035

4

.075

.0150

200

.870

.0013

5

.090

.0139

300

.950

.00046

10

.150

.0105

400

.980

.00020

15

.190

.0089

500

.994

.00006

20

.240

.0080

700

.9992

.00000

30

.310

.0070

1000

.99995

.00000

Thus it appears that 90 per cent of the upward flux is confined to
a coastal belt about 250 miles wide. Finally the stream line equation
(188) becomes

17*l = C2 (199)

_ ,93e-01* — .

_ e-?rcos H 1

for the San Diego region, and the stream lines corresponding to C2
equal 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 are shown in
figure 17.

•15

0

liOO

Fig. 17. Traces of theoretical surfaces of flow on a vertical plane perpen-
dicular to the coast. The flow included between any two consecutive surfaces
is one-tenth of the total amount.

406 MISCELLANEOUS STUDIES

Figure 17 represents the component of the hypothetical circulation
in a plane perpendicular to the coast corresponding to a uniform \\ind
over the whole region, in which the bottom bends sharply upward at
the coast, but gives some idea of the actual circulation. In examining
the figure it must be noted that the vertical is very much greater than
the horizontal scale. In fact, if the horizontal scale were the same
as the vertical one actually used the diagram would be about one and
one-half miles in length.

DEDUCTION OF THE UPWELLING VELOCITY OFF SAN DIEGO FROM THE
OBSERVED RELATION OF SALINITY TO DEPTH

If the rates of molecular diffusion of salts and conduction of heat
are relatively very small as compared with the rate of transfer due to
the alternating circulation (p. 368), the differential equation

W (200)

6t dy2 dy

(equation 80, p. 368) applies in general where the constant p- is a
measure of the rate of transfer, and the dependent variable is the salt
concentration or temperature. An application to temperature data
has already been made, and we have only to replace 0 by the salinity
8 in the temperature equation and its solution already worked out
(pp. 375-378) in order to obtain the corresponding formulae for
salinity. However, the salinity data are too incomplete to furnish
reliable estimates of averages for each month, and it seemed best to
use the data taken in the same region (Section 404) for each of the
three months, August, 1912. February, 1913, and April, 1913, which
correspond to an interval of less than one year. These data are
presented in table 15.

OCEAN TEMPEEATUEES

407

TABLE 15

Observed Salinities for August, 1912, February, 1913, and April, 1913,

in Section 40t

Depth

Salinities

y

(i = 2)
February

0-4)

April

0-8)

August

Mean annual
values

0

33.47

33.58

33.75

33.61

10

33.47

33.58

33.69

33.58

20

33.44

33.58

33.58

33.51

30

33.42

33.61

33.54

33.48

40

33.46

33.66

33.57

33.52

50

33.50

33.73

33.65

33.57

60

33.54

33.79

33.73

33.63

70

33.57

33.85

33.57

33.67

80

33.62

33.90

33.81

33.72

90

33.66

33.94

33.85

33.76

100

33.70

33.99

33.88

33.79

150

34.03

34.19

34.07

34.05

200

34.25

34.30

34.17

34.21

300

34.32

34.33

34.23

34.27

400

34.33

34.34

34.28

34.30

500

34.36

34.35

34.32

34.34

600

34.40

800

34.48

1100

34.53

From the formula for salinity (replacing & by 8 in equation 117)
it follows that the mean of any two salinities corresponding to a time
interval of six months would equal approximately the mean annual
salinity. Accordingly, the means of the salinities for February and
August are assumed to be the mean annual salinities in this case. The
constants C, D, and A of equation (117) were found as in the case
of temperature data by fitting the equation

Sm=Ce** + D (201)

to the observed mean annual salinities, Sm. Then the observed mean
annual salinity was subtracted from each of the entries under February
and April, and the expression

was subtracted from each of these, using the same numerical values
for wt, r and a as on page 377, and the value of C determined above.
The remaining expression

Me™ cos (at -f &!# — e')

408

MISCELLANEOUS STUDIES

(p. 377) was fitted to these remainders. The values of the constants
thus found are D = 34.55, C = — 1.25, A= —.005, M=— .6,
f' = — 65, a1==— .0075, /i2= 10,600, wj==— 53, 61==— .00495. The
value of &! determined from equation (121) exceeds numerically the
value — .00175 determined from the observations. Expressing the
angle in degrees, these values are — .28 and — .10 respectively. The
values n2 = 10,600 and wl = — 53 obtained from the salinity data are
in good agreement with the values 7760 and — 31 obtained from the
more complete and extensive temperature data (p. 378). The com-
puted and observed values of the remainders and of the mean annual
salinities are entered in table 16 as an additional test of the theory.
The computed remainders were obtained from

_ .6 e-. DOTS, cos (30£ _ ly _j_ 65) °

and the computed mean annual salinities were obtained from

34.55 — 1.25 e-005".

TABLE 16

Computed and Observed Remainders for February and April, and the Computed
and Observed Mean Annual Salinites

Remainders

Mean annual

Depth

Salinities

< = 2 Feb.

t = 4 April

y

Computed

Observed

Computed

Observed

Computed

Observed

50

.14

-.07

.32

.16

33.53

33.57

100

.07

-.09

.21

.20

33.74

33.79

150

.02

-.02

.13

.14

33.96

34.05

200

.00

.04

.08

.09

34.04

34.21

300

-.01

.05

.03

.06

34.22

34.27

400

-.01

.03

.01

.04

34.33

34.30

500

-.01

.02

.00

.01

34.40

34.34

600

34.44

34.40

800

34.48

34.48

1100

34.50

34 . 53

The mean velocity of upwelling can also be estimated from the
salinity distribution in the upper 30 meter layer, and by an entirely
different method. A comparison of this value with the two estimates
made with the aid of theoretical results already presented affords a
severe test of the theories and gives an idea of the reliability of the

OCEAN TEMPEEATUBES

409

Water Sc/rface

results. In dealing with mean annual values we can assume all con-
ditions to be independent of the time, from which it follows that the
total amount of water in a given volume remains constant and the
total amount of salts remain constant. Therefore the rate of flow of

water and salts into the volume must
equal the rate of flow out of the
volume.

This principle will be applied to
two different volumes, thus giving two
estimates of the velocity of the up-
welling. First, consider a vertical
column (fig. 18) whose cross section
is a square of unit area and whose
base is at the depth y2 where the
salinity has its minimum value (Mc-
Ewen, 1916, p. 272).

The explanation of symbols used
follows :

Coast

Bottom

S = the salinity at any depth y.
$0 = the salinity at the surface.
S2 = the salinity at the depth y2.
r2Wl = the vertical velocity at the

depth y2, Wi is the maximum

value, and corresponds to large

values of y (table 13).
y = the horizontal velocity.
E = ihe rate of evaporation at the

surface.

Fig. 18. Eectangular volume of
water from the depth y2 of mini-
mum salinity to the surface, used
in determining the velocity of up-
welling from salinity.

A flow into the volume is regarded
as negative, and a flow out is regarded
as positive, the vertical distances and
velocities are regarded as positive
when directed downward from the surface, and horizontal distances
and velocities are positive when directed away from the coast.
Because of the invariability of the amount of water in the volume

r2W1 — E+fvdA =

(202)

410 MISCELLANEOUS STUDIES

where dA is an element of the vertical surface enclosing the column,
and the integral is taken over the whole vertical surface. Similarly
because of the invariability of the amount of salts in the volume

r2W18a +fvSdA = 0. (203)

Let

S = S + *S (204)

where S is the constant mean salinity for the whole volume and AS
is a variable increment. Then equation (203) becomes

(205)

and substituting the value of J Vd A from equation (202) we have
raW18a — '8(r.tWl — E) +JV(Afl)<Li=() (206)

or solving for W1

— HE— Cv(±8)dA

W,= - J- - = - (207)

r2(82 — S)

An estimate of I V(&S)dA can be made as follows: Let the

volume be so turned that two of its parallel faces are parallel to the
coast line and therefore perpendicular to the horizontal velocity T
directed away from the coast and given by equation (p. 401 )

(208)

Then neglecting the variation of the salinity in a direction parallel to
the coast, the integral

y (209)

Cv(*8)dA=— Cvi(*S)1dy + Cv
J J 0 J ,

where Vl and (A$)j correspond to the face next to the coast and V2
and (AS) 2 to the face farthest from the coast. From a study of our
salinity observations (McEwen, 1916, especially plates 20, 21, 22. and
24) made from five to fifteen miles offshore, it appears that the hori-
zontal gradient parallel to the coast is negligible as compared to that
perpendicular to the coast, thus justifying equation (209). The
numerical values of the horizontal salinity gradient per meter esti-
mated from our observations are given for a series of depths in table 17.

OCEAN TEMPEEATVEES

41]

TABLE 17

Mean horizontal salinity gradient per meter during the summer for a series

of depths

Depth, y

0

5

10

15

20

25

30

Salinity
gradient ,A$'

6X10-6

6X10-6

3X10-6

0

-10-6

-3X10-6

-6X10-6

Our salinity data indicate that (A$') is practically zero in winter,
hence the mean annual value would be about half of that entered in
the table.

Owing to the small value of (Fx — V2) compared to the mean
value V and because A$' = (A$)2 — (A/S)j equation (209) can be
written in the form

(210)

(211)

//*2/2
V(±S)dA= I T
•ft

From equations (207, 208, and 210) we have finally

— SE-VJ(z) I (Ls")e-«i/cos fj— ^1 d2/
W =- Jo _ V 7

r.Ofif.— fl)

In order to check the above results the same principle will be
applied to a different volume (fig. 19). The stream lines (fig. 17)
being traces of surfaces of flow on a plane perpendicular to the coast,
two such planes, two surfaces of flow, and two horizontal planes
inclose a volume such that the component of the velocity along a line
parallel to the coast is the same for each vertical plane. Hence the
vertical flux through a horizontal section of this volume must be
independent of the depth of the section, in order that the total quantity
of water inclosed by these surfaces may be constant. Consider the
volume inclosed by two surfaces of flow, two vertical faces perpen-
dicular to the coast and parallel to the plane of the paper at unit
distance apart (fig. 19), and two horizontal sections at the depths yz
and y1 of which the upper forms the base of a rectangular prism
extending upward to the surface of the water.

Let rlW1 be the mean vertical velocity at the depth yl, and r,W1
that at the depth y2, then rlW1B1 must equal r2W^B2 where B, nnd
B2 are the areas of the upper and lower sections respectively, whence
the section areas Bl and B2 must satisfy the equation

B,

(214)

412

MISCELLANEOUS STUDIES

For the whole volume inclosed, from the base B2 to the water
surface, the condition of the constancy of the quantity of water re-
quires that

r2W1B2 — EB^ CvtdA4 — Cv3dA3 = 0 (215)

Coast:

Wdf~c7 Sen face

Bottom

Fig. 19. Volume of water included in part by two surfaces of flow from the
depth j/2 of minimum salinity to a depth ylt used in determining the velocity
of upwelling from salinity.

where (dA3) is an element of area of the vertical face of the prism
next to the coast, (dA4) is an element of the other parallel face, and
V3 and V4 are the corresponding horizontal velocities.

Similarly, in order that the total amount of salts may remain
constant,

r2W^B2 + fv4S4dA4 — Cv3S3dA3 = 0 (216)

where $3 and $4 are the salinities corresponding to the elements of
area (dA3) and (dAt).

OCEAN TEMPEBATUBES 413

Let

flf.^ + AS,

and

84 = l3i + ±St (217)

where $! is the constant mean salinity for the prismatic volume from
the surface to the depth y^ and A$3 and A$4 are variable increments.
Then equation (216) becomes

r.W^A + 8 A fVtdAt — fvsdA3 I

0, (218)

and substituting the value of

from equation (215) we have

W1 = ^— — 1« =— (219)

where (AS') is defined on page 411, and finally, substituting for V
the value given by equation (208)

V0f(z)

j (Aflf'

r1(82 — 8l) rl(S2—S1) (220)

If the depth of the upper section is at the level y2 we must sub-
stitute j/2 for yt and r2 for rx in equation (220), which then becomes
identical with equation (211). But in equation (220) yl can have
any value between the limits, zero and j/2, where y2 is the depth of
minimum salinity, and estimates of the velocity based on different
values of y1 should give the same result. Some divergence of these
values in any actual case is to be expected, since the different esti-
mates are based on different observations that are subject to errors
of measurement and since the actual relation of the velocity to depth
may differ from the theoretical relation (p. 403).

In table 18, where V0 = 324000 (p. 404) and — E for the latitude
of San Diego is .0754 meters per month (Schmidt, 1915, p. 121) are
presented the results based upon the mean value of (AS'), that is,
half the value entered in table 17, the mean annual salinities as shown
by plate 11 and table 3 (McEwen, 1916) and the values of rx and
/(10) from tables 13 and 14.

414

MISCELLANEOUS STUDIES

"5M^J*O>OOOOW^»flW

fe~

^^ l^ ^J^ CO CO CO CO CO ^* *O

7111111111

/*yi

osot-ooTjiosooe^oo

Jo ' (* "I"

O 00 CO OO CO CO CO Is- ^t* *O
5lCN(N(M<NlNO<(NtNCN

riOS,-ST)

/-»2/i
i /•* \

.05 1 (AS'Je-^cos 1 -r -ay } dy

^O 00 00 Is* t"^ Is* C*5 CO CO C^

i^ c^ o^ *o O5 05 oo oo ^r oo

1 \4 /

fc/O

i— 1 i— 1 i-H i— 1 t-H

J2/i
(•»>-„(,-,)*

00^(<OOi-lC35O5COCOOOCO

COCNt^'-ii-I^Hi-HOOOCO
i-ii-HC^(M(NCN(N'-H^

oa MS

TfCN^OJOOt-CNt-t-.O

1 ^

CO O5 l^* CD *O CD CO CD CD l"*»

7111111111

B,

00 C^ **^ CO O5 00 OO t*" ts* CO

I

IC'O'OOIOOIC'O'O'O
<N C< <N CN (N CN (N <N (N CN

C

COCOi-Hi— lOOOOt^-tfSlOOO

?— ( -^ co t^» co o oo o co i~^

*~H C^ OO ^ O CO 1^* 00 00 OS

W

I

oo
oO'-iOoor>-coic<*'3|oo

^-(i-H^HOOOOOOO

1 1 1 1 1 1 1 1 1 1

loo

>OiOOOiOO500O^Q>-(
^CNOO500t>-t>-COCOiO
COCOCOiOU3U3»OiC>OiO

co oo co co co co oo co co oo

00 OO 00 00 OO OO OO CO 00 00

s

OO«OOCNiOOO'-i'«tit--O
i-H r-t i-H C1 CN CN OO

OCEAN TEMPERATURES 415

Except for the first two values of y^ near the surface, where the
water is most disturbed, and for the largest value of t/15 where the
difference (S2 — SJ is so small that it is subject to the largest pro-
portional error, the computed values of the velocity Wx shown in the
last column are in good agreement. And the mean of the central
values, which are subject to the least error, is about — 35, which is
the best estimate from the available data and agrees well with the
values — 31 and — 53 found before, page 408.

CONCLUSION

In the case of no average flow of the water the form of a function
giving the rate of gain of heat and of another giving the rate of loss
of heat from a small volume of water at a given latitude and depth,
was developed from a few simple assumptions, suggested by laboratory
experiments as well as field observations. Equating the sum of these
two expressions to the product of the specific heat by the volume by
the rate of change of temperature resulted in a differential equation
whose solution gave the temperature from the surface to a depth of
ten meters as a known function of time, depth, latitude, and certain
physical constants, under the conditions of no average flow of water.

From observations on the relation to latitude of the mean annual
surface temperature and the annual temperature range and the relation
to latitude of the mean annual solar radiation and its annual range, all
of the physical constants of the formula were computed.

The lag between the time of the temperature maxima and minima
and the time of the maximum and minimum values of the solar radia-
tion deduced from these constants agreed well with the observed value.
Also the mean monthly temperature at a region whose mean annual
temperatures agree well with the normal value for the latitude, that
is, the value corresponding to no average flow, would be expected to
agree with those computed from the formula for normal temperature.
A comparison of such computed and observed temperatures for a
series of latitudes from 20° N to 40° N indicated a very satisfactory
agreement.

If the rate of change of temperature due to some factor not in-
cluded in the above reasoning is known this quantity can be added
to the differential equation already derived. Since all of the constants
of the original differential equation are known the solution of the
modified equation will give the temperature due to the new factor.

416 MISCELLANEOUS STUDIES

If, for example, there is a flow of the water, the rate of flow multiplied
by the temperature where the water enters a given element of volume
gives the rate at which the heat is carried into the volume. The rate
at which the heat leaves the volume, computed in the same way,
subtracted from the rate at which it enters gives the rate of change
of heat in the volume due to the corresponding ocean current, whether
horizontal or vertical. A term expressing this rate of change of heat
in the case of a horizontal current was added to the differential
equation, and the solution furnished a means of estimating the magni-
tude of horizontal currents from surface temperatures without con-
sidering the causes of the currents.

Numerical applications of this formula to a region of the North
Pacific off the California coast and of the North Atlantic off the
African coast gave estimates of the horizontal flow in good agreement
with direct observations and with what would be expected from the
observed wind velocity.

Conclusive evidence of the presence of currents directed upward
from the bottom along the California ccast which cause reduction in
temperature has been published before; but to test this conclusion
further it was assumed that the reduction of the temperature of the
coastal water was due entirely to the upwelling of deep water, and
the temperature distribution at depths exceeding 40 meters was
assumed to result from a flow of heat according to Fourier's well
known conductivity equation, in which a term expressing the rate of
loss of heat due to upwelling was added. The formal solution of this
equation contained certain physical constants whose evaluation re-
quired the observed monthly temperatures at a series of depths. Our
temperature data for a deep water region twenty miles offshore from
San Diego were sufficient for making approximate estimates of all of
the constants, of which the velocity of upwelling and the term corre-
sponding to conductivity are of special interest. The latter constant
depends largely upon the eddy motion, or alternating circulation,
which tends to mix the water and has been called eddy conductivity or
Mischungsintensitat, and was found in this case to be several thousand
times the laboratory value. In dealing with salinities the same
formula can be used; and a similar constant appears, which also
depends largely on the mixing motion of the water, and would be
expected to have practically the same value as that determined from
temperatures. Our salinitiy data, though not so complete as the
temperature data, confirmed this conclusion and gave approximately

417

the same velocity of upwelling. Moreover, in applying hydrodynamical
equations to problems of oceanic circulation the coefficient of viscosity
must be replaced by another constant depending on the eddy, or
turbulent motion, and having a much greater value than the laboratory
value obtained from observations on a slow laminar flow free from
irregular motions. Similar results have also been found by G. I.
Taylor in certain recent studies of the temperature, water vapor, and
velocity in the atmosphere. The results of laboratory experiments
and theories based on them were helpful but could not provide the
numerical values required; in each case field observations were neces-
sary. Furthermore, since the eddy conductivity, or Mischungs-
intensitat, is not a physical constant of the substance, sea water or
air, but depends upon the intensity and character of the circulation,
its value will vary accordingly. The following approximate values of
these constants, the coefficient of viscosity, diffusion, and conductivity
under laboratory conditions and estimated from field observations in
the ocean and the atmosphere, illustrate the great differences between
field and laboratory conditions.

TABLE 19

Estimates of the coefficients of viscosity, diffusion, and heat conductivity made

from field observations in the ocean and atmosphere compared

ivith values obtained in laboratory experiments

Sea Water

Observer

Coefficient

Laboratory value
in c. g. a. units

Value from field
observations in
c. g. s. units

Ratio of the field
to the laboratory
value

Ekman
Jacobsen

Viscosity1
Diffusion

.014
.0000125

217
.3 to 11.4

15,500
24,000 to

McEwen

Diffusion

.0000125

40

320,000

McEwen

Conductivity2

.0012

30

25,000

Air

Taylor
Taylor

Viscosity1
Conductivity2

.13

.20

770 to 6,900
570 to 3,400

6,000 to 50,000
3,000 to 17,000

1 The laboratory value of the "kinetic coefficient" of viscosity, or the coefficient of viscosity
divided by the density, is given since it is the constant in the equations of motion, which is
formally equivalent to the one given by field observations.

2 The laboratory value of the thermometric conductivity is given since that is the constant
in the equation of heat conductivity, which is formally equivalent to the one given by field
observations.

The estimation of the effect of upwelling on the surface tempera-
ture made necessary the consideration of results obtained for depths

418 MISCELLANEOUS STUDIES

below the 40 meter level and the solution of the original differential
equation for surface temperatures after adding a term giving the rate
of temperature change due to upwelling. The monthly values deduced
in this way for the San Diego region agree very well with those
afforded by the observations.

From the magnitude of the vertical velocity found from tempera-
tures and from certain results deduced from Ekman 's hydrodynamical
theory, the distribution of the horizontal and vertical velocity of the
water in a vertical plane perpendicular to the coast was deduced and
represented graphically.

An independent estimate of the velocity of upwelling made from
the distribution of salinities in the upper 30 meter layer and of the
rate of evaporation at the surface agreed well with the other two
estimates. Moreover, the estimates of the velocity of upwelling from
the temperature or salinity distribution did not depend upon the
cause of the upwelling ; but it is an interesting fact that such a vertical
current would be expected along the California coast from Ekman 's
hydrodynamical theory.

I wish to express my obligation to Dr. W. E. Ritter of this institu-
tion, and to my laboratory assistant, Mr. Nephi W. Cummings, for
his aid in making the computations and for his suggestions while
preparing the manuscript.

Transmitted June 26, 1918.
Scripps Institution for Biological Research
of the University of California,
La Jolla, California.

OCEAN TEMPERATURES 419

LITERATUEE CITED
BJERKNES, V. F. K.

1901. Cirkulation relative zu der Erde. Ofversikt af Kongl. Vet.-Akad.

Forhandl., 58, 739-757.
EKMAN, V. W.

1905-06. On the influence of the earth's rotation on ocean currents. Arkiv
for Matematik, Astronomi och Fysik, 2, 1-53, 1 pi. and 10 figs.

1906. Beitrage zur Theorie der Meerestromungen. Ann. d. Hydrog. u.

marit. Meteor., 34, 423-430, 472-484, 527-540, 566-583, 38 figs.

GEHBKE, JOHAN.

1909. Beitrag zur Hydrographie des finnischen Meerbusens. Finnlandische

hydrographisch-biologische Untersuchungen, III. (Leipzig, Engel-
mann), pp. 1-39, pis. 1-3, 7 figs, in text.

1910. Beitrage zur Hydrographie des Ostseebassins. Cons, perman. explor.

Mer., Circ. 52, 1-195, 2 pis.

GREIN, K.

1913. Untersuchungen iiber die Absorption des Lichts in Seewasser. I.
Die selektive Absorption des Sonnenlichts im Meer nebst einer
Beschreibung des Instruments und der Methode. Ann. Inst.
Oceanogr., 5, fasc. 6, pp. 1-23, 7 figs, in text.
HANN, JULIUS.

1915. Lehrbuch der Meteorologie. (Leipzig, Tauchnitz), xiv + 847 pp., 28

pis., 69 figs, in text.
HELLAND-HANSEN, B.

1911-12. The ocean waters, an introduction to physical oceanography. T,
General part (methods). Intern. Rev. d. ges. Hydrobiol. u. Hydrog.,
3, Hydrog. Suppl., pp. 1-84, figs. 1^6 in text.
JACOBSEN, J. P.

1913. Beitrag zur Hydrographie der danischen Gewasser. Medd. Komm.
Havandersogelser (Hydrografi), 1, no. 2, 94 pp., 14 pis., 17 figs, in
text.
KRUMMEL, O.

1907. Handbuch der Ozeanographie. (Stuttgart, Engelhorn), 1, xv + 526

pp., 60 figs, in text.

1911. Handbuch der Ozeanographie. (Stuttgart, Engelhorn), 2, xvi -j- 766

pp., 182 figs, in text.
KRUMMEL, O., AND RUPPIN, E.

1905. Ueber die innere Reibung des Seewasser. Wiss. Meeresuntersuch-

ungen, N. F., 9, 29-36.
KNOTT, C. G.

1903-05. Ocean temperatures and solar radiation. Proc. Royal Soc. Edin-
burgh, 25, part 1, pp. 173-184.
MAKAROFF, S.

1894. Le "Vitiaz" et 1'ocean Pacifique. (St. Petersburg, Kummar), 1,
xliii + 337 pp.

420 MISCELLANEOUS STUDIES

McEwEN, G. F.

1912. The distribution of ocean temperatures along the west coast of North
America deduced from Ekman 's theory of the upwelling of the
cold water from the adjacent ocean depths. Intern. Eev. d. ges.
Hydrobiol. u. Hydrog., 5, 243-286, 21 figs, in text.

1914. Peculiarities of the California climate explained on the basis of

general principles of atmospheric and oceanic circulation. U. S.
Mon. Weather Rev., 42, 14-23, 13 figs.

1916. Summary and interpretation of the hydrographie observations made
by the Scripps Institution for Biological Research of the Univer-
sity of California, 1908-1915. Univ. Calif. Publ. Zool., 15, 255-356,
pis. 1-38.

MICHAEL, E. L., AND McEwEN, G. F.

1915-16. Hydrographie, plankton and dredging records of the Scripps Insti-
tution for Biological Research of the University of California,
1901-12. Univ. Calif. Publ. Zool., 15, 1-206, 4 figs, and 1 map in
text; ibid., 16, 207-254, 4 figs, in text.

MOHN, H.

1887. The Norwegian North Atlantic expedition, 1876-1878: The North
Sea, its depths, temperature and circulation. (Christiania, Gro%-
dahl), 212 pp., 48 pis.

MOORE, W. I.

19'08-11. Meteorological charts of the North Pacific Ocean. U. S. Weather
Bureau, Washington, D. C.

MURRAY, Sir JOHN.

1898. On the annual range of temperature in the surface waters of the ocean
and its relation to other oceanographic phenomena. Geog. Jour.,
12, 113-137.

MURRAY, Sir JOHN, AND HJORT, JOHAN.

1912. The depths of the ocean. (London, Macmillan), xx + 821 pp., 9 pis.,

575 figs, in text.
NANSEN, F.

1913. The waters of the north-eastern North Atlantic. Intern. Rev. d. ges.

Hydrobiol. u. Hydrog., 4, 1-139, pi. 17, figs. 1-52 in text.

PETERSEN, JOHANNES

1912. Hydrographie und Meteorologie Finnlands und der benachbarten
Meere, nach "Atlas de Finlande." Ann. Hydr. u. marit. Meteor.,
40, 131-145, 4 figs, in text.

PULS, CASAR.

1895. Oberflachentemperaturen und Stromungsverhaltnisse des Aequatorial-
giirtels des Stillen Ozeans. Aus dem Archiv der deutschen See-
warte, 18, 1-38, pis. 1-4.

SCHMIDT, WILHELM.

1915. Strahlung und Verdunstung an freien Wasser flachen. Ein Beitrag zum

Warmehaushalt des Weltmeers und um WTasserhaushalt der Erde.
Ann. Hydr. u. marit. Meteor., 43, 111-124.

OCEAN TEMPERATURES 421

SCHOTT, G.

1895. Die jahrliche Temperaturschwankung des Oceanwassers. Petermanns

Mittelungen, 41, 153-159, pi. 10.
1902. Oceanographie und maritime Meteorologie. Wissenschaftliche Ergeb-

nisse der deutschen Tief see-Expedition auf dem Dampfer "Val-

divia," 1898-1899. (Jena, Fischer), 1, xii + 403 pp., 26 pis. and

35 figs in text.
SCHOTT, GERHARD, UND SCHU, FRITZ.

1910. Die Warmeverteilung in den Stillen Ozeans. Ann. Hydr. u. marit.

Meteor., 38, 1-26, 15 pis. 1 fig. in text.

SCHOTT, G., SCHULZ, B.r UND PERLEWITZ, P.

1914. Die Forschungsreise S.M.S. " Mb'we, " im Jahre 1911. Aus dem Archiv

der deutschen Seewarte, 37, vi + 104 pp., 8 pis. and 16 figs, in text.

TAYLOR, G. I.

1915. Eddy motion in 'the atmosphere. Philos. Trans. Eoyal Soc. London,.

215, ser. A, pp. 1-26, figs. 1-5 in text.

THORADE, H.

19'09. Ueber die Kalifornische Meeresstromungen, Oberflachentemperaturen
und Stromungen an der Westkiiste Nordamerikas. Ann. Hydr. u.
9 marit. Meteor., 37, 17-34, 63-77, pis. 5, 10, 11, 5 figs, in text.

• 1914. Die Geschwindigkeit von Triftsstromungen und die Ekmansche
Theorie. Ibid., 42, 379-391, 3 figs, in text.

WEGEMANN, G.

1905rt. Die vertikale Temperaturverteilung im Weltmeere durch Warmeleit-
ung. Wissenschaftliche Meeresuntersuchungen, herausgegeben von
der Kommission zur Untersuchung der deutschen Meere in Kiel
und der Biologischen Anstalt auf Helgoland. Abteilung Kiel,
N. F., 8, 137-143.

.1905&. Ursachen der vertikalen Temperaturverteilung im Weltmeere unter
besonderer Beriicksichtigung der Warmeleitung. Ann. Hydr. u.
marit. Meteor., 33, 206-211.

WHARTON, W. J. L.

1894. Presidential address, Section E, Geography. Eept. Brit. Assoc. Adv.
Sci.

WlNKELMANN, A.

1906. Handbuch der Physik. (Leipzig, Earth), 3, xiv + 1178, 206 figs, in
text.

ZOPPRITZ, K.

1878. Hydrodynamische Probleme in Beziehung zur Theorie der Meeres-
stromungen. Wiedemanns Annalen, 3, 582-607.

CHANGES IN THE .
CHEMICAL COMPOSITION
OF GRAPES DURING RIPENING

BY

F. T. BIOLETTI, W. V. CRUESS, AND H. DAVI

[University of California Publications in Agricultural Sciences, Vol. 3, No. 6, pp. 103-130]

CHANGES IN THE

CHEMICAL COMPOSITION OF GRAPES
DURING RIPENING

BY

F. T. BIOLETTI, W. V. CRUESS, AND H. DAVI

The investigations reported in this paper were undertaken to
determine the changes in chemical composition of vinifera varieties
of grapes in California during the growing and ripening stages. A
survey of the literature indicated that, although the subject had been
quite fully investigated in Europe with vinifera varieties and in
America with the native varieties, very little had been published upon
the ripening of vinifera varieties under California Conditions. A
great many analyses of different varieties of grapes have been made
by chemists of the University of California Experiment Station, nota-
bly by G. E. Colby, and are reported in the publications of this station.1
A paper by G. E. Colby2 gives data upon the nitrogen content of a
number of varieties of ripe vinifera grapes. Most of the analyses,
however, do not show the changes in composition during ripening.

Of the more recent European investigations3 some deal with the
changes in general composition, others are confined to a discussion of a
single component, such as sugar, or coloring matter, or acid principles.

The changes in composition of American varieties of grapes during
ripening have been studied quite thoroughly by W. B. Alwood4 and
his associates. These investigations gave particular attention to the

1 Hilgard, E. W., The composition and classification of grapes, musts, and
wines. Eept. of Viticultural Work, Univ. Calif. Exper. Sta. Rep., 1887-93, pp.
3-360.

2 Colby, G. E., On the quantities of nitrogenous matters contained in Cali-
fornia musts and wines. Ibid., pp. 422-446.

s Kelhofer, W., The grape in the various stages of maturity; trans, by E.
Zardetti. Gior. Vin. Ital., vol. 34 (1908), no. 30, pp. 475-477.

Barberon, G., and Changeant, F., Investigations on the development and

[103]

424 MISCELLANEOUS STUDIES

increase in sugar content and changes in acidity during the period
in which the grapes were under observation. Alwood and other mem-
bers of the Bureau of Chemistry, United States Department of Agri-
culture, have also published a number of reports4 on the general
composition of American varieties of grapes as affected by season,
locality, etc.

The most notable changes taking place during ripening were found
by the European and American investigators mentioned above to be :
(1) increase in total sugar ; (2) decrease in ratio of glucose to fructose ;
(3) decrease in total acid; (4) increase in ratio of cream of tartar to
total acid due to decrease in total acid ; (5) decrease in tannin ; and (6)
increase in coloring matter. The cream of tartar and protein change
very little in percentage during ripening, although, according to the

composition of varieties of grapes in Abraon-Durso. Ann. Soc. Agr Sci. et Ind.,
Lyon (8), vol. 1 (1903), pp. 97-159.

Laborde, J., The transformation of the coloring matter of grapes during
ripening. C. R. Acad. Sci. (1908), vol. 17, pp. 753-755.

Martinand, V., On the occurrence of sucrose and saccharose in different parts
of the grape. C. R. Acad. Sci. (1907), vol. 24, pp. 1376-79.

Eoos, L., and Hughes, E., The sugar of the grape during ripening. Ann.
Falsif. (1910), vol. Ill, p. 395.

Bouffard, A., Observations in regard to the proportion of sugar during ripen-
ing. Ann. Falsif. (1910), vol. Ill, pp. 394-5.

Zeissig, Investigations on the process of ripening on one-year-old grape wood.
Ber. k. Lehranst. Wien, Obst-u. Garten-bau (1902), pp. 59-64.

Koressi, F., Biological investigations of the ripening of the wood of the
grape. Rev. Gen. Bot., vol. 13 (1901), no. 149, pp. 193-211; no. 150, pp. 251-264;
no. 151, pp. 307-325.

Brunet, R., Analysis and composition of the grape during ripening. Rev.
de Viticulture, vol. 37, pp. 15-20.

Garina, C., Variations in the principal acids of grape juice during the process
of maturing. Canina. Ann. R. acad. d'agricultura di Torino, vol. 57 (1914),
p. 233. Cf. Ann. Chim. applicata, vol. 5 (1914), pp. 65-6. See also Ann. r. acad.
d'agr. di Torino, vol. 57, pp. 233-90.

Baragolia, W. L, and Godet, C., Analytical chemical investigations on the
ripening of grapes and the formation of wine from them. Landw. Jahrb., vol.
47 (1914), pp. 249-302.

Riviere, G., and Bailhache, G., Accumulation of sugar and decrease of acid
in grapes. Chem. Abs. Jour. (1912), p. 1022; Jour. Soc. Nat. Hort. France (4),
pp. 125-7; Bot. Cent., 1912, pp. 117, 431.

Pantanelli, Enzyme in must of overripe grapes. Chem. Abs. Jour., vol. VI
(1912), p. 2447.

< Alwood, W. B., Hartmann, J. B., Eoff, J. R., and Sherwood, S. F., Develop-
ment of sugar and acid in grapes during ripening. U. S. Dept. Agric. Bull. 335,
April 11, 1916.

The occurrence of sucrose in grapes. Jour. Indust., vol. II, Eng. Chem.

(1910), pp. 481-82.

Sugar and acid content of American native grapes. 8th Inter. Cong.

Appl. Chem. (1912), Sect. Vla-XIv, pp. 33, 34.

Enological Studies: the chemical composition of American grapes grown

in Ohio, New York, and Virginia. U. S. Dept. Agric. Bur. Chem. Bull. 145, 1911.
— Crystallization of cream of tartar in the fruit of grapes. U. S. Dept.
Agric. Jour. Agric. Research (1914), pp. 513, 514.

Alwood, W. B., Hartmann, B. G., Eoff, J. R., Sherwood, S. F., Carrero, J. O.,
and Harding, T. J., The chemical composition of American grapes grown in the
central and eastern states. U . S. Dept. Agric. ( 1916) Bull. 452.

CHEMICAL COMPOSITION OF GRAPES 425

investigations referred to, there is a slight increase in both of these
constituents.

In the investigations reported in the present paper, particular
attention was given to increase in total solids and sugar, decrease in
total acid, and changes in protein and cream of tartar in the must or
juice of the grapes. The ripening of the leaves was traced by noting
the changes in starch, sugar, acid, and protein content.

Sampling. — During 1914 and 1915 samples of fruit were taken
from the time the grapes had reached full size but were still hard and
green until they had become overripe. During 1916 the first samples
were taken shortly after the berries had set and before the seeds had
formed. The last samples were taken when the grapes had become
overripe. Samples of leaves were also taken in 1916 on the same
dates that samplings of the grapes were made. The samples were
taken at intervals of approximately one week. They were in all cases
taken from the experimental vineyard at Davis.5

Five-pound samples of grapes were used. The grapes were picked
from the first crop, except in 1914, when a comparison of the ripening
of first and second crops was made. An ordinary five-pound grape
basket was filled with leaves at each sampling. The samples of grapes
and leaves were shipped from the vineyard to the laboratory at
Berkeley, where the grapes were placed in an Enterprise fruit crusher
and pressed. The juice was sterilized in bottles at 212° F. The leaves
were ground in an Enterprise food chopper and sterilized at 212° F
in wide mouth, air tight bottles. The samples were then reserved for
chemical examination.

In 1914 it was found that there was considerable irregularity in
the variation of samples from week to week. For example, instead
of an increase of total solids during the periods between samplings, a
slight decrease was found in a few samples. During the 1915 season
it was therefore considered of interest to note what effect certain
factors might have upon the composition of samples taken on the
same date.

1. Effect of Age of Vine. The entire first crop from three large
old vines and from three small young vines, all of the Muscat variety,
was picked, crushed, and pressed. Analyses of the juices were made
with the following results :

5 The authors wish to express their appreciation of the assistance of F. C.
Flossfeder, of the University Farm at Davis, who gathered most of the samples
reported upon in this paper.

1105}

426 MISCELLANEOUS STUDIES

TABLE 1 — EFFECT OF AGE OF VINE ON BALLING AND ACID OF MUST OF MUSCAT

GRAPES

Vine Balling Acid

Small, no. 1 24.7 .67

Small, no. 2 27.7 .49

Small, no. 3 27.6 .67

Large, no. 1 22.0 .88

Large, no. 2 23.5 .75

Large, no. 3 23.6 .76

Average, small 26.7 .61

Average, large 23.0 .81

Difference 3.7 — .20

The results show rather strikingly that young vines ripen their
fruit earlier than do mature vines. This fact makes it essential that
samples, to be comparative, must be taken from vines of the same age.

2. Comparison of Grapes from North and South Sides of Vines.
The whole first crop from three large Muscat vines was picked. The
bunches from the north and south sides of each vine were kept sep-
arate. They were crushed, pressed, and analyzed for Balling and acid
content.

TABLE 2 — COMPARISON OF BALLING AND Aero OF JUICE FROM GRAPES PICKED FROM
NORTH AND SOUTH SIDES OF VINES

Vine and side of vine Balling Acid

1-N 21.3 .92

1-S 22.7 .84

2-N 23.5 .81

2-S 23.5 .80

3-N 23.1 .81

3-S 24.1 .71

Average, N side 22.63 .85

Average, S side 23.43 .78

Difference 80 —.07

The tests indicate that grapes located on the south side of the vine
ripen more rapidly than those on the north side. This difference is
apparently due to the fact that the south side of the vine receives
more heat than the north side.

3. Effect of Location of Bunch on Cane. Grapes of first crop, from
canes showing two bunches each, were picked and the bunches from
near the bases of the canes kept separate from those near the tip of
the cane. They were crushed, pressed, and analyzed for Balling and
acid.

[106]

X

CHEMICAL COMPOSITION OF GRAPES 427

TABLE 3 — EFFECT OF LOCATION OF BUNCH ON CANE

Nearest base of cane Nearest tip of cane

Vine Balling Acid Balling Acid

Muscat, no. 1, cane 1 25.1 .73 23.7 .83

Muscat, no. 1, cane 2 25.6 .79 24.8 .80

Muscat, no. 2, cane 1 25.1 .85 24.6 .87

Muscat, no. 2, cane 2 25.2 .78 24.7 • .85

Muscat, no. 3, cane 1 23.0 .79 22.6 .82

Muscat, no. 3, cane 2 24.5 .73 23.8 .73

Muscat, no. 4, cane 1 24.2 .90 25.2 .90

Muscat, no. 4, cane 2 24.5 .68 23.8 .83

Tokay, cane 1 21.2 .67 21.2 .80

Tokay, cane 2 23.0 .63 22.4 .76

Sultanina, cane 1 23.3 .61 22.3 .62

Sultanina, cane 2 22.5 .61 23.0 .63

Sultana, cane 1 23.2 .78 21.6 .70

Sultana, cane 2 21.1 .90 20.0 1.20

Palomino, cane 1 25.1 23.5

Palomino, cane 2 22.0 23.7

Means 24.9 .75 23.1 .81

The data indicate that bunches at the base of the cane ripen in most
cases more rapidly than those near the tip, although this relation does
not always hold and may be reversed in some instances.

4. Variation in Balling Degree of Must from Bunches of Similar
Appearance and Size from Same Vineyard and Gathered on Same
Date. A five-pound basket of grapes of first crop and selected for
similarity of color, size of bunch, and general appearance was picked
from each of a number of vines in the same vineyard. Vines of
similar size and appearance were chosen. Several varieties were rep-
resented in the experiment. Tests of Balling degree only were made.

TABLE 4 — VARIATION IN BALLING IN MUST FROM GRAPES OF SAME VARIETY PICKED
FROM DIFFERENT VINES OF SIMILAR APPEARANCE

Vine Mean Maximum

Variety number Balling Balling variation

Cornichon 3 14.5 ......

Cornichon 6 15.0

Cornichon 9 14.2

Cornichon 11 14.7

Cornichon 16.1 14.9 1.9

Emperor 10 12.0

Emperor 11 14.5

Emperor 13 15.2

Emperor 14 15.5

Emperor 17 15.0 14.4 3.5

Malaga 5 . 18.5

Malaga 6 17.2

[107]

MISCELLANEOUS STUDIES

TABLE 4 — (Continued)

Vine Mean Maximum

Variety number Balling Balling variation

Malaga 7 19.7

Malaga 9 18.5

Malaga 11 19.2 18.6 2.0

Muscat * 21.7

Muscat * 21.1

Muscat * 20.9 ....:.

Muscat * 21.5

Muscat * 21.7 21.4 .8

Palomino 3 19.5

Palomino 4 21.0

Palomino 6 21.2

Palomino 7 20.7

Palomino 9 18.8 20.2 2.4

Sultanina * 22.5

Sultanina * 21.5

Sultanina * 18.7

Sultanina * 22.0

Sultanina * 22.6 21.5 3.9

Tokay * 19.8

Tokay * 19.3

Tokay * 18.7

Tokay * 20.7

Tokay * 19.5 19.6 2.0

Pedro Zumbon 7 21.5

Pedro Zumbon 4 21.2

Pedro Zumbon 6 20.6

Pedro Zumbon 3 18.5

Pedro Zumbon 5 19.8 20.3 3.0

Emperor 15 18.1

Emperor 8 15.8

Emperor 14 16.2

Emperor 9 16.8

Emperor 16 16.3 16.6 2.3

Cornichon 4 17.3

Cornichon 9 16.3

Cornichon 10 17.9

Cornichon 11 17.8

Cornichon 13 18.0 17.5 1.7

Malaga 4 18.3

Malaga 5 20.4

Malaga 6 20.0

Malaga 8 20.1 19.7 1.8

Mean variation, six ripest varieties 2.32

Mean variation, six least ripe varieties 2.20

Average variation, whole series 2.30

Adjacent vines.

CHEMICAL COMPOSITION OF GRAPES 429

The data illustrate the difficulty of selecting five-pound lots of the
same variety that will represent average samples.

5. Effect of Location of Berries on the Bunch. All of the bunches
of the first crop were taken from two Muscat vines. • The bunches
were cut into top and bottom halves. These lots were crushed sep-
arately, pressed, and the juices analyzed.

TABLE 5 — EFFECT OF LOCATION OF BERRIES ON BUNCH

Sample Balling Acid

Vine no. 1, stem end of bunch 23.6 .76

Vine no. 1, apical end of bunch 22.7 .87

Vine no. 2, stem end of bunch 21.3 .92

Vine no. 2, apical end of bunch 21.3 .93

The results show that considerable variation in composition of the
berries may exist within the same bunch.

6. Effect of Thoroughness of Pressing. About ten pounds of Mus-
cat grapes were crushed and lightly pressed. The pulp and skins left
from this pressing were then thoroughly crushed and pressed a second
time. The juices from the two lots were analyzed separately.

TABLE 6 — EFFECT OF THOROUGHNESS OF PRESSING

Sample Balling Acid

First pressing 22.8 .78

Second pressing 22.8 .79

There was practically no difference between the juices from lightly
and thoroughly pressed grapes of the same lot.

The data from the above six tests indicate that it is a very difficult
matter to select grapes that will represent a fair average sample of
the grapes to be studied. The size and age of the vine, the side of
the vines, the location of the bunch on the cane, and individual vines,
all affect the composition of the juice from the grapes very materially,
and these factors should be taken into account when samples are
taken.

Preservation of Samples and Preparation for Analysis. — In 1914
the samples of juice were preserved with HgCl2, 1 :1000. In 1915 and
1916 the samples were sterilized at 100° C. Before analysis the bottles
were heated to 100° C for an hour to dissolve any cream of tartar which
might have separated. The juices were filtered before analysis. Con-
siderable coagulation of dissolved solids took place during sterilization.

[1091

430 MISCELLANEOUS STUDIES

Methods of Analysis. — The samples were analyzed by the methods
in use in the Agricultural Chemistry Laboratory and the Nutrition
Laboratory of this station. A brief description of the methods follows :

1. Total Solids. The juice was filtered clear and cooled below
15° C. The specific gravity was determined by a pycnometer at
15? 5 C. The corresponding total solids, or extract, was found from
Windisch's tables in Leach's Food Analysis, page 697. This table
gives the extract as "grams per 100 grams"; that is, per cent by
weight. To calculate the corresponding grams per 100 c.c., the per
cent by weight was multiplied by the specific gravity. This gives a
figure not very much greater than grams per 100 grams in juices of
low specific gravity, but gives a figure as much as 2 per cent greater
where the total solids are much above 20 per cent. The two methods
of reporting total solids has in the past led to much unnecessary
confusion. It is therefore urged that the reader bear in mind the
distinction between the two methods when reading the discussions in
this paper or examining the curves.

2. Sugar. The sample was filtered; an aliquot was treated with
lead acetate ; diluted to mark ; filtered ; lead removed with anhydrous
Na2C03, and the sugar determined in an aliquot by the gravimetric
method, using Soxhlet's modification of Fehling's solution. The Cu20
was weighed directly after drying at 100° C. The corresponding
sugar as invert sugar was obtained from Munson and Walker's table
in Leach's Food Analysis. The grams of invert sugar per 100 c.c.
found in this way was divided by the specific gravity of the must to
obtain the corresponding grams per 100 grams of juice.

3. Total acid was determined by titration of a 10 c.c. sample with
N/10 NaOH, using phenolphthalein as an indicator, and is reported
as tartaric acid, grams per 100 c.c.

4. Cream of tartar was estimated by a method suggested by Pro-
fessor D. R. Hoagland of the Division of Agricultural Chemistry.
Ten c.c. of the juice was incinerated at a low heat in a muffle furnace
until well carbonized, but not to a white ash. (Excessive heating
results in loss of K by volatilization.) The K2C03 formed by incin-
eration was leached out with hot water and a known excess of N/10
HC1 added. This was titrated back with N/10 NaOH, using methyl
orange as an indicator. The K2C03 is obtained by difference and
calculated back to cream of tartar, assuming that all of the K2C03 is
formed by the oxidation of cream of tartar, KH(C4H406). It is

[110]

CHEMICAL COMPOSITION OF GRAPES 431

reported as grams KH(C4H4O6) per 100 c.c., and also as tartaric
acid.

5. Free Tartaric Acid was obtained by difference between total acid
and cream of tartar calculated as tartaric acid. It is reported as
grams per 100 c.c.

6. Protein in the juice was determined by the usual Kjeldahl-
Gunning method upon a 10 c.c. sample. It is reported as grams per
100 c.c.

7. Moisture in the leaves was determined by drying the sample at
100° C.

8. Sugar in the leaves was estimated by leaching the dried sample
with cold water and determining sugar by the gravimetric Fehling
method in the nitrate.

9. Starch in the leaves was determined by hydrolysis of the dried
ground sample with dilute HC1 at 100° C., followed by nitration and
the usual gravimetric Fehling method for juice described above.

10. Protein in the leaves was determined by the Kjeldahl-Gunning
method on .5 gram samples.

11. Acid in the leaves was estimated by leaching in hot water and
titrating in the presence of the leaves, using litmus paper as indicator.

Analyses of Musts from Grape-Ripening Samples, 1914, 1915, 1916.
The data from the analyses have been assembled in the following
tables. Owing to the size of the tables, abbreviations have been
necessary for the headings of the columns.

EXPLANATIONS OP HEADINGS OF TABLES

1. Sp. gr. = Specific gravity at 15?5 C.

2. T. S. G. = Total solids in grams per 100 grams.

3. T. S. C. = Total solids in grams per 100 c.c.

4. S.G, = Sugar in grams per 100 c.c.

5. S. T. = Sugar in grams per 100 grams.

6. Tl. A. = Total acid in grams per 100 c.c.

7. C. T. = Cream of tartar in grams per 100 c.c.

8. C. T. T. = Cream of tartar as tartaric acid, grams per 100 c.c.

9. T. A. = Total free acid as tartaric obtained by subtracting cream of tartar
as tartaric from total acid as tartaric.

10. P. = Protein, grams per 100 c.c.

11. S. = Sum of sugar, cream of tartar, tartaric acid, and protein in grams
per 100 c.c.

12. T. S. — S. = Total solids (T. S. C.) — S (preceding column).

[nil

432

MISCELLANEOUS STUDIES

TABLE 7 — GEAPE RIPENING TESTS, 1914
(Grapes from Davis)

Malaga

First crop:

Variety

l

2

3

4

5

6 7

8

9

10

11

12

and date

8p.gr.

T. 8. O.

T. 8. C.

8. G.

S.I.

Tl. A. C. T.

C. T. T.

T. A.

P.

8.

T. 8. 6.

Aug. 19

1.0396

10.25

10.65

7.32

7.04

2.78 .35

.13

2.65

.21

10.53

.12

Aug 26

1.0413

10.69

11.13

7.84

7.53

2.65 .36

.14

2.51

.25

10.96

.17

Aug. 26

1.0595

15.42

16.33

13.37

12.62

.77 .48

.19

.58

.55

14.98

1.41

Aug. 26

1.0613

15.87

16.84

14.31

13.50

1.46 .31

.12

1.34

.33

16.29

.55

Aug. 26

1.0694

18.01

19.25

16.59

15.52

1.00 .36

.14

.86

.38

18.19

1.06

Aug. 31

1.0732

19.00

20.39

17.65

16.45

.87 .55

.22

.65

.45

19.30

1.09

Sept. 23

1.0736

19.10

20.50

17.83

16.60

.74 .38

.15

.59

.52

19.32

1.18

Oct. 5

1.0965

25.12

27.54

24.89

22.70

.72 .50

.20

.52

.57

26.48

1.06

Second crop

Aug. 10

1.0213

5.51

5.62

2.07

2.03

3.22 .23

.09

3.13

.17

5.60

.02

Aug. 31

1.0495

12.82

13.45

9.58

9.13

2.51 .40

.16

2.35

.28

12.61

.84

Sept. 14

1.0532

13.78

14.51

11.89

11.30

2.07 .37

.15

1.92

.31

14.49

.02

Sept. 23

1.0670

17.43

18.60

15.29

14.33

1.54 .50

.20

1.35

.29

17.43

1.17

Sept. 23

1.0869

22.59

24.55

22.04

20.19

1.07 .45

.18

.89

.41

23.79

.76

Oct. 5

1.0930

24.20

26.45

23.90

21.87

.94 .48

.19

.75

.41

25.54

.91

Tokay

(

First crop:

Aug. 2

1.0454

11.75

12.28

8.73

8.35

2.63 .46

.18

2.45

.32

11.96

.32

Aug. 10

1.0624

16.08

17.08

14.28

13.44

1.56 .45

.18

1.38

.27

16.38

.70

Aug. 19

1.0682

17.69

18.90

15.94

14.92

1.32 .45

.18

1.14

.27

17.80

1.10

Aug. 3i

1.0849

22.09

23.97

21.87

20.16

.63 .59

.23

.40

.40

23.26

.71

Sept. 4

1.0865

22.49

24.44

22.21

20.44

.77 .43

.17

.60

.32

23.56

.88

Sept, 4

1.0912

23.72

25.88

23.44

21.48

.59 .64

.25

.44

.41

24.93

.95

Sept. 23

1.0937

24.38

26.66

24.15

22.08

.58 .49

.19

.30

.39

25.33

1.33

Oct. 14

1.0991

25.80

28.36

25.55

23.25

.45 .54

.21

.24

.45

26.78

1.58

Oct. 14

1.1000

26.04

28.64

25.78

23.44

.52 .58

.23

.29

.58

27.23

1.41

Second crop

Aug. 19

1.0657

17.04

18.16

15.03

14.10

1.91 .50

.20

1.70

.32

16.55

.61

Sept. 14

1.0701

18.19

19.47

16.68

15.59

1.29 .52

.21

1.11

.33

18.64

.83

Sept. 23

1.0769

19.95

21.48

19.22

17.85

1.01 .48

.19

.82

.40

20.92

.56

Oct. 14

1.0911

23.70

25.86

23.43

21.47

.69 .60

.24

.45

.40

24.88

.98

TABLE 8 — GRAPE RIPENING TESTS,

1915

(Grapes from

Davis)

Cornichon

Variety

1

2

3

4

5

6 7

8

9

10

11

12

and date

Sp.gr.

T. S. O.

T. 8. C.

8.0.

8. I. Tl. A. C. T.

C. T. T.

T. A.

p.

8.

T. 8. S.

Aug. 22

1.0324

8.38

8.65

3.99

3.86

3.05 .58

.23

2.82

.38

7.77

.88

Sept. 1

1.0514

13.31

13.99

10.70

10.18

1.62 .61

.25

1.37

.42

13.10

.89

Sept. 15

1.0688

17.85

19.08

15.94

14.91

.97 .70

.28

.69

.43

17.76

1.32

Sept. 22

1.0723

18.76

20.12

16.97

15.83

.94 .71

.28

.66

.46

18.80

1.32

Sept. 29

1.0737

19.13

20.54

18.31

17.05

.87 .75

.30

.61

.66

20.33

.21

Oct. 7

1.0781

20.28

21.86

19.41

18.02

.71 .73

.29

.42

.48

21.04

.82

Oct. 14

1.0843

21.91

23.76

20.40

18.81

.78 .68

.27

.62

.66

22.36

1.40

Oct. 22

1.0873

22.70

24.68

21.06

19.37

.75 .78

.31

.44

.46

22.74

1.94

[112]

CHEMICAL COMPOSITION OF GEAPES

433

TABLE 8 — (Continued)

Emperor

Variety

l

2

3

4

5

6

7

8

9

10

ll

12

and date

Sp. gr.

T. S. G.

T. S. C.

S. G.

S.I.

Tl. A.

C. T.

C.T.T.

T. A.

p.

S. T.

A. S.

Aug. 19

1.0420

10.87

11.33

6.96

6.68

2.33

.38

.15

2.18

.38

9.90

1.43

Sept. 1

1.0479

12.40

12.99

9.82

9.37

1.89

.40

.16

1.73

.62

12.57

.42

Sept. 7

1.0560

14.51

15.32

11.48

10.87

1.70

.47

.19

1.57

.54

14.00

1.32

Sept. 15

1.0632

16.37

17.40

14.88

14.00

1.40

.53

.21

1.18

.54

17.13

.27

Sept 22

1.0652

16.91

18.01

15.46

14.51

.93

.48

.19

.74

.55

17.23

.78

Sept. 29

1.0672

17.43

18.60

16.37

15.34

.91

.48

.19

.72

.66

18.23

.37

Oct. 7

1.0744

19.31

20.75

17.82

16.59

.79

.58

.23

.56

.51

19.47

1.28

Oct. 14

1.0765

19.86

21.38

18.37

17.06

.79

.59

.24

.56

.63

20.15

1.23

Oct. 22

1.0792

20.57

22.20

19.81

18.36

.75

.63

.25

.49

.66

21.59

.61

Malaga

Aug. 19

1.0546

14.14

14.91

12.47

11.82

2.05

.36

.15

1.90

.75

15.48

.57

Aug. 2n

1.0651

16.86

17.96

14.53

13.64

1.66

.46

.18

1.48

.90

17.37

.59

Sept. 1

1.0678

17.59

18.78

16.75

15.69

1.38

.44

.18

1.20

.89

19.28

.50

Sept. 7

1.0719

18.66

19.50

17.00

15.86

1.29

.44

.18

1.11

.70

19.25

.25

Sept. 15

1.0758

19.68

21.17

18.17

16.89

1.21

.62

.25

.96

.70

20.45

.72

Sept. 22

1.0760

19.81

21.32

18.39

17.09

1.18

.61

.25

.93

.74

20.67

.65

Sept. 29

1.0812

21.20

22.92

18.48

17.09

1.07

.58

.23

.84

.75

20.65

2.27

Oct. 7

1.0838

21.78

23.61

21.03

19.40

1.07

.65

.26

.81

.73

23.22

.39

Oct. 14

1.0970

25.25

27.70

24.58

22.41

.59

.83

.33

.26

.88

26.55

1.15

Muscat

Aug. 19

1.0615

15.94

16.92

13.93

13.12

1.70

.36

.15

1.55

.70

16.54

.38

Aug. 25

1.0744

19.31

20.75

17.96

16.72

1.21

.62

.25

.96

.62

20.16

.59

Sept. 1

1.0805

20.91

22.59

19.50

18.05

.79

.63

.25

.54

.63

21.30

1.29

Sept. 7

1.0827

21.47

23.25

20.39

18.83

.76

.65

.26

.50

.66

22.20

1.05

Sept. 15

1.0917

23.85

26.04

23.49

21.52

.96

.58

.23

.73

.58

25.38

.66

Sept. 22

1.0954

24.14

26.44

24.54

22.40

.77

.62

.25

.52

.85

26.53

.09

Sept. 29

1.1048

27.30

30.16

27.01

24.45

.72

.72

.29

.44

.72

28.89

1.27

Oct. 7

1.1079

28.12

31.15

28.28

25.53

.66

.59

.23

.43

.66

29.96

1.19

Pedro Zumbon

Aug. 19

1.0555

14.38

15.18

11.96

11.33

1.81

.68

.27

1.54

.33

14.51

.67

Aug. 25

1.0588

15.24

16.14

13.77

13.01

1.09

.57

.23

.86

.53

15.73

.41

Sept. 1

1.0642

16.64

17.71

15.61

14.67

.58

.52

.21

.37

.43

16.93

.78

Sept. 7

1.0693

17.98

19.23

16.55

15.48

.84

.48

.19

.65

.73

18.41

.82

Sept. 15

1.0708

18.37

19.67

18.17

16.97

.56

.58

.23

.33

.64

19.72 .

.05

Sept. 22

1.0912

23.72

25.88

23.02

21.10

.53

.87

.35

.19

.64

24.72

1.16

Sultana

Aug. 19

1.0673

17.80

19.00

15.63

16.64

1.69

.33

.13

1.56

.32

17.84

1.16

Aug. 25

1.0746

19.37

20.82

17.96

16.71

1.44

.37

.14

1.30

.38

20.01

.81

Sept. 1

1.0815

21.17

22.90

20.26

18.73

1.14

.54

.22

.92

.50

22.22

.68

Sept. 7

1.0893

23.22

25.29

23.02

21.13

.78

.44

.18

.60

.34

24.40

.89

Sept. 22

1.0902

23.39

25.50

23.10

21.19

1.24

.50

.20

1.04

.38

25.02

.48

Sept. 29

1.0922

23.99

26.20

. 24.04

22.01

.80

.41

.17

.63

.42

25.50

.70

[113]

434

MISCELLANEOUS STUDIES

TABLE 8 — (Continued)

Sultanina

Variety

1

2

3

4

5

6

7

8

9 10

11

12

and date

Sp. gr.

T. S.Q

. T. S. 0.

S. G.

S.I.

Tl. A.

C. T.

C.T.T.

T. A. P.

8.

T. A. b.

Aug. 19

1.0673

17.46

18.64

15.87

14.87

1.27

.44

.18

1.09 .42

17.82

.82

Aug. 25

1.0743

19.26

20.69

18.30

17.03

1.19

.47

.19

1.00 .37

20.14

.55

Sept. 1

1.0771

20.02

21.56

18.98

17.62

.85

.49

.20

.65 .42

20.54

1.02

Sept. 7

1.0892

23.20

25.27

22.42

20.58

.72

.80

.32

.40 .62

24.24

1.03

Sept. 15

1.0927

24.12

26.36

23.62

21.62

.79

.76

.30

.39 .45

25.22

1.14

Sept. 22

1.0984

25.62

28.14

25.71

23.41

.60

.58

.23

.37 .45

27.11

1.03

Sept. 29

1.1049

27.33

30.20

27.41

24.81

.54

.51

.20

.34 .42

28.68

1.52

Toicay

Aug. 19

1.0598

15.50

16.43

14.41

13.60

1.74

.41

.16

1.58 .29

16.69

.26

Aug. 25

1.0676

17.54

18.73

15.63

14.64

1.24

.39

.15

1.09 .69

17.80

.93

Sept. 1

1.0757

19.65

21.14

18.17

16.89

.84

.47

.19

.66 .44

19.74

1.40

Sept. 7

1.0781

20.28

21.86

19.11

17.73

.79

.45

.18

.61 .37

20.54

1.32

Sept. 15

1.0785

20.39

21.99

19.26

17.86

.74

.48

.19

.55 .40

20.69

1.30

Sept. 22

1.0798

20.73

22.38

20.17

18.68

.59

.51

.20

.39 .36

21.43

.95

Sept. 29

1.0823

21.38

23.14

20.76

19.18

.85

.58

.23

.62 .28

22.24

.90

Oct. 7

1.0830

21.57

23.36

20.87

19.27

.69

.63

.25

.44 .42

22.36

1.00

Oct. 14

1.0851

22.12

24.00

21.53

19.84

.65

.69

.28

.38 .36

22.96

1.04

Oct. 22

1.0895

23.28

25.36

22.91

21.03

.66

.72

.29

.37 .37

24.37

.99

TABLE 9 — GRAPE EIPENING TESTS,

1916

Burger

Variety

1

2

3

4

5

6

7

8

9 10

11

12

and date

Sp.gr.

T. S. G. T. S. C.

S. G.

S. I.

Tl. A.

C. T.

C. T. T.

T. A. P.

8.

T. S. S.

June 12

1.0212

5.48

5.59

1.29

1.55

2.95

.55

.22

2.73 .44

5.27

.32

June 19

1.0195

5.04

5.88

.87

.88

2.88

.51

.21

2.67 .45

4.51

1.37

June 27

1.0220

5.69

5.82

1.25

1.28

2.94

.33

.13

2.81 .45

4.87

.95

July 7

1.0220

5.69

5.82

1.11

1.28

2.98

.49

.20

2.78 .31

4.86

.96

July 10

1.0200

5.17

5.27

.93

.95

3.32

.57

.23

3.09 .37

4.97

.30

July 19

1.0205

5.30

5.41

1.03

1.05

3.13

.55

.22

2.91 .35

4.86

.55

July 27

1.0225

5.82

5.95

1.13

1.15

2.93

.48

.19

2.74 .34

4.71

1.24

Aug. 3

1.0258

6.67

6.84

2.14

2.19

2.71

.63

.25

2.46 .40

5.68

1.16

Aug. 7

1.0330

8.53

8.83

3.36

3.46

2.67

.87

.35

2.32 .47

7.12

1.21

Aug. 16

1.0391

10.11

10.51

5.90.

6.13

.2.41

.95

.38

2.03 .46

9.57

.94

Aug. 23

1.0422

10.92

11.38

6.03

6.27

2.10

.98

.39

1.71 .63

9.59

1.89

Aug. 30

1.0529

13.70

14.42

9.95

10.42

1.15

1.03

.41

.74 .49

12.70

1.72

Sept. 5

1.0645

16.73

17.81

14.51

15.43

1.01

1.07

.43

.68 .61

17.79

.02

Sept. 12

1.0717

18.61

19.94

16.27

17.36

.95

.98

.39

.56 .82

19.72

.22

Sept. 20

1.0765

19.86

21.37

17.44

18.73

.87

1.06

.42

.45 .62

20.86

.51

Sept. 26

1.0808

20.99

22.68

18.48

19.99

.81

1.01

.40

.41 .83

22.24

.44

Cornichon

June 12

1.0202

5.22

5.32

.91

.93

3.15

.64

.26

2.89 .32

4.78

.54

June 19

1.0200

5.17

5.27

.86

.88

2.96

.62

.25

2.71 .42

4.63

.64

June 27

1.0193

4.99

5.08

.84

.86

2.89

.39

.16

2.73 .56

4.54

.44

July 7

1.0201

5.19

5.29

.87

.89

2.88

.44

.18

2.70 .52

4.55

.74

July 10

1.0206

5.32

5.43

.85

.87

3.27

.54

.22

3.05 .53

4.99

.44

July 19

1.0225

5.82

5.95

1.28

1.30

3.11

.57

.23

2.88 .55

5.30

.65

July 27

1.0242

6.25

6.40

1.63

1.66

2.94

.54

.22

2.72 .44

5.26

.14

Aug. 3

1.0373

9.65

10.00

5.00

5.19

2.87

.59

.24

2.63 .56

8.97

1.03

[114]

CHEMICAL COMPOSITION OF GSAPES

435

TABLE 9 — (Continued)

. Variety

l

2

3

4

5

6

7

8

9

10

11

12

and date

Sp. gr.

T. S. G.

T. S. C.

S. G.

S.I.

Tl. A.

C. T. C

T.T.

T. A.

P.

S.

T. A. S.

Aug. 7

1.0375

9.70

10.06

5.28

5.48

2.79

.65

.26

2.53

.66

9.32

.64

Aug. 16

1.0434

11.23

11.71

6.30

6.57

2.75

1.06

.43

2.32

.53

10.48

1.23

Aug. 23

1.0635

16.47

17.51

12.19

12.96

1.85

1.06

.43

1.42

.58

16.02

1.49

Aug. 30

1.0685

17.77

18.97

14.75

15.61

1.16

1.10

.44

.72

.63

18.06

.91

Sept. 5

1.0694

18.01

19.25

15.03

16.07

.93

.90

.36

.57

.58

18.12

1.13

Sept. 12

1.0757

19.65

21.09

16.37

17.60

.87

1.14

.46

.41

.78

19.97

1.12

Sept. 20

1.0786

20.41

22.00

17.52

18.88

.84

.94

.37

.44

.59

20.85

1.15

Sept. 26

1.0828

21.52

23.30

18.52

20.03

.72

.83

.33

.39

.85

22.10

1.20

Muscat

June 12

1.0203

5.25

5.35

.91

.93

2.93

. .65

.26

2.71

.38

4.67

.68

June 19

1.0199

5.14

5.24

.70

.72

3.37

.63

.25

3.12

.44

4.91

.33

June 27

1.0210

5.43

5.54

1.33

1.36

3.33

.48

.19

3.14

.49

5.47

.07

July 7

1.0210

5.43

5.54

1.63

1.66

3.32

.54

.22

3.10

.45

5.75

.21

July 10

1.0195

5.04

5.14

1.33

1.36

3.60

.55

.22

3.38

.36

5.65

.51

July 19

1.0251

6.49

6.65

2.55

2.61

3.40

.58

.23

3.17

.49

6.85

.20

July 27

1.0308

7.97

8.22

3.56

3.67

2.67

.66

.26

2.01

.45

6.79

1.43

Aug. 3

1.0488

12.64

13.26

9.72

10.19

1.77

.68

.27

1.50

.46

12.83

.43

Aug. 7

1.0582

15.68

16.58

12.72

13.53

1.60

.73

.29

1.31

.55

16.12

.46

Aug. 16

1.0803

20.86

22.53

16.81

18.15

1.16

.94

.38

.78

.51

20.38

1.70

Aug. 23

1.0910

23.67

25.82

20.20

22.04

.82

1.04

.42

.40

.56

24.04

1.78

Aug. 30

1.0972

25.30

27.75

21.87

22.99

.65

1.21

.49

.16

.58

25.94

1.81

Sept. 5

1.1023

26.64

29.36

23.28

24.74

.60

1.17

.47

.13

.65

26.69

2.67

Sept. 12

1.1101

28.70

31.85

25.95

27.83

.56

1.35

.54

.02

.69

29.89

1.96

Sept. 20

1.1122

29.25

32.72

26.43

29.39

.68

1.56

.63

.05

.58

31.27

1.45

Sept. 26

1.1133

29.54

32.89

26.68

29.70

.56

1.39

.56

.00

.59

31.57

1.32

TABLE 10 — CATAWBA GRAPE RIPENING TESTS

(Table from U. S. Dept. Agric
Catawba

, Bulletin 335, by W. B.

Alwood)

1912:

. '

Variety

l

o

3

4

5

6

7

8

9

and date

Sp. gr.

T. S. G.

T. S. C.

S.I.

S. G.

Tl. A.

C. T.

C. T. T.

Days

Sept. 4

1.0329

8.51

8.84

3.60

3.72

3.68

.39

.16

0

Sept. 9

1.0419

10.84

11.29

6.68

6.96

3.02

.41

.16

5

Sept. 12

1.0515

13.34

14.03

9.35

9.78

2.48

.46

.18

8

Sept. 17

1.0537

13.91

14.66

10.38

10.95

2.12

.45

.18

13

Sept. 24

1.0569

14.74

15.58

11.33

11.96

1.74

.53

.21

20

Oct. 1

1.0614

15.92

16.89

12.75

13.48

1.63

.54

.22

27

Oct. 7

1.0663

17.20

18.34

13.79

14.71

1.53

.61

.24

33

Oct. 16

1.0725

18.82

20.18

15.35

16.46

1.34

.61

.24

42

Oct. 23

1.0716

18.58

19.90

15.01

16.09

1.28

.59

.24

47

Oct. 29

1.0769

19.97

21.50

16.49

17.75

1.22

.57

.23

53

Nov. 4

1.0790

20.52

22.14

16.77

18.08

1.28

.71

.28

59

Nov. 8

1.0755

19.60

21.07

16.39

17.61

1.09

.52

.21

63

nisi

436 MISCELLANEOUS STUDIES

Curves of Total Solids, Sugar, Total Acid, Free Acid, and Cream
of Tartar. — In order to present the data in a form in which they may
be readily studied, graphs have been constructed using time in days
as abscissae and the above constituents expressed in grams per 100 c.c.
as ordinates. The curves represent the data for 1914, 1915, and 1916.
For comparison, curves of the changes in composition of Catawba
grapes reported by W. B. Alwood in the United States Department
of Agriculture Bulletin 335 have been included. The acid principles
have been plotted to a scale five times as great as that used for total
solids and sugar in order that the variations in acidity might be more
apparent.

Discussion of Graphs of Total Solids, Sugar, Total Acid, Cream of
Tartar, and Free Acid. — (1) Total Solids and Sugar. The data are
more complete for 1916 than for 1914 or 1915, and include the period
during which the berries are growing to full size as well as the ripen-
ing period itself, during which the rapid increase in sugar occurs.
The curves for 1916, therefore, are of more interest than those for
1914 and 1915. In the case of the Burger variety, total solids and
sugar remained constant for approximately forty days after the tests
were started. There was then a slight rise in these components for
a period of about ten days. From that point on the rise in total solids
and sugar was very rapid and fairly uniform. The behavior of the
Cornichon was very similar.

The Muscat began ripening about ten days earlier than the Burger
and Cornichon, and proceeded much more rapidly up to about the
ninetieth day after the experiment was started. There was then a
slowing up in the increase in total solids and sugar corresponding to
the period of over-ripeness. This slower increase in total solids is
also evident in the curves for Emperor, Muscat, Sultana, and Tokay
for the 1915 season, and would undoubtedly show in all cases if the
observations were continued sufficiently.

The effect of the season upon the rate of ripening is shown by a
comparison of the Cornichon and Muscat varieties for 1915 and 1916.
All varieties ripened more slowly in 1915 than in 1916, resulting in
steeper curves for 1916. However, owing to the fact that sampling
was started later in 1914 and 1915 than in 1916, the curves for the
former two years show only the changes taking place during the latter
half of the ripening period. No very close comparisons therefore can
be made of the three years.

The Catawba reported by Alwood, and for which curves appear

[116]

CHEMICAL COMPOSITION OF GBAPES

437

CROP*

9/4-

x.

&

lartanc.

G<

-Jfn.i

XX,

Cream efTgffa.

^ /o

E.V

3t

/Or/ari'c

Tar far

IQOCJL.

2+

£C+_

13 JtO

Fig. 1 — Malaga first and second crops, 1914.

[117]

438

MISCELLANEOUS STUDIES

nihlfa4r«*

/ ht

aad-Lj-eaaL

XV-

&^

ZO&.

15

ft

(A

10 5

V

m of Taffor

Fig. 2 — Tokay first and second crops, 1914.

[118]

CHEMICAL COMPOSITION OF GBAPES

439

COrfN/CjJON

/? = Tfitnlflr.J

Tc/rfa

fe: an of

of

d ^in-tar <3rr?3

/OOd-

ft
&0

18
16

\\

s

IX 13 #4 50 36
TIMC IN OfF/5

60 66

Fig. 3 — Cornichon and Emperor, 1915.

[119]

440

MISCELLANEOUS STUDIES

i

Kb

a

*7.

it

K)&

6 7x

^F7 30 55

IN

Fig. 4 — Malaga and Muscat, 1915.

[120]

s

CHEMICAL COMPOSITION OF GKAPES

441

to*.

to K /<3 &f 30 36 -f/C" -73

tO 66

"-Total,

\UL7QNft

of ~fertir Gm

Per /dO

tt

2C4-

«L.

It

10

&
6

_&&

e /*> TO #*
T/M£ IN D/?y S

Fig. 5 — Pedro Zuinbon and Sultana, 1915.

[121]

442

MISCELLANEOUS STUDIES

~IZ 73 T& 70 3T

Ifll

Fig. 6 — Sultanina and Tokay, 1915.

66

[122]

CHEMICAL COMPOSITION OF GSAPES

443

TIME IN DftYS

Fig. 7 — Burger and Cornichon, 1916.

[123]

444

MISCELLANEOUS STUDIES

JO "iv

TIME IN o/rys

Fig. 8— Muscat, 1916.

Fig. 9— Catawba (U. S. Dept. Agric. Bull. 335).

[1241

CHEMICAL COMPOSITION OF GRAPES 445

in figure 9, ripened more slowly than the Vinifera varieties. For
example, during a period of fifty days, the total solids increased only
4 per cent. It can not be said from the data at hand whether this
slow ripening is due to the conditions under which the grapes were
grown or to the variety.

By reference to figures 1 and 2 it may be seen that the general
form of the ripening curves is the same for the first and for second
crop. In one case, the Malaga, the curves are almost identical for
the period common to both, i.e., from 10.6 Bal. to 26.3 Bal., showing
an equal rate of ripening. In the other, the Tokay, the curve of the
second crop, from 18.2 Bal. to 24.6 Bal., is much flatter than that of
the first, indicating a rate of ripening with the latter of about two
and a half times that of the former. This difference can be accounted
for by the cooler weather during the time the second crop Tokay was
ripening, which was about ten days later than in the case of the second
crop Malaga. The slower ripening is probably due both to the direct
effect of the cool weather and to the decreased activity of the leaves
at lower temperatures.

(2) Changes in Total Acid, Cream of Tartar, and Free Acid.
Owing to the fact that the analyses were started in 1914 and 1915
after ripening had commenced, the curves for these years show a
decrease in acid throughout the period of the tests. In 1916, however,
a rise in total acid occurred during the growing stage, as shown by
a rise in the curve during the first thirty days of the experiment.
Although this rise is not very large, it is quite definite, and occurs
in all three varieties tested. The rise was most positive in the case
of the Muscat grape, and amounted to .67 per cent acid as tartaric.
From the point of maximum acidity, the total decreases slowly until the
period of rapid ripening sets in. The total acid then decreases very
rapidly for a time and more or less in proportion to the increase in
total solids and sugar. As the grapes near maturity, the rate of de-
crease of total acid becomes less and the total remains practically
constant after the grapes have reached maturity.

The cream of tartar in general increases very slightly during the
periods of growth and ripening.

The increase in total acid during the first stages of growth is due
to increase in the free acid. Since the cream of tartar remains almost
constant throughout the ripening period, the curve of the free acid
is practically parallel with that of the total acid.

As the grapes approach maturity, the cream of tartar calculated as

[1251

446

MISCELLANEOUS STUDIES

tartaric acid approaches the total acid, and in one case, (Musct, 1916),
actually became equal to the total acid, indicating that in this instance
no free acid remained.

Second crop grapes were found to be higher in free acid than
first crop grapes of the same total solids and sugar content. The
Catawba grape grown under eastern conditions (fig. 9) exhibits rela-
tively high free acid. Alwood6 has found this free acidity in eastern
grapes to be due largely to malic acid. No attempt was made in the
analyses of the California samples to identify the various acids making
up the free acidity which was calculated as tartaric acid.

Mean Differences Between Total Solids and Sugar. — The following
table contains figures representing the differences between total solids
and sugar at the various percentages of total solids indicated at the
tops of the columns. The data represent a range of total solids from
5 per cent to 30 per cent. The figures were taken from the data
reported in tables 7 to 9, and represent several varieties of grapes.
Only a few determinations of total solids and sugar were available
for the lower concentrations (5 per cent to 15 per cent), and therefore
the figures for this range may not represent averages so accurately
as the figures above 15 per cent total solids.

Between 5 per cent and 11 per cent solids, the average difference
between total solids and sugar remains practically constant. From
11 per cent to 17 per cent total solids, the mean difference decreases
quite rapidly. From 17 per cent to 30 per cent, the difference remains
fairly constant. The variations noted after 17 per cent total solids

Fig. 10 — Mean differences between total solids and sugar between 5 per cent
and 30 per cent total solids.

e U. S. Dept. Agric. Bull. 335.

[126]

CHEMICAL COMPOSITION OF GSAPES 447

t> CO

to £ co "2. ::::!: If., to to
o> "S 2-

.*• " : : : :::::: ?° !*• !^
co co ::::i:i::cocn~3

co

en K-» :::::cn

co to
#.

co eo

s

os g p

CO *• CO ^ g

b to :;i:i:i!l:booM

I I

to co «o to en w n> &

H

co co co eo i_i <->• 45

b\ co

CO ......... COCOfcOMg£l

J-» co, : i : i i i i i i bo io co w =: g

" w

to . ...... toeoeototoi-'p g

bo en i i i i i i i «o rf»- co bo Ic-

oo 1

on OT

to .... to co to to to co co to M *

o> QO to b ^ co ^ co b w •* » W

to i w

to ..... tO to tO tO CO CO tO HI ? ^

bo ~j iiiiiio-vjbo-vicocoto00

rf^. ^ M

P ^

to ,_, . cotocotototocotoeototoi-' 2.

bo H-> ocoto«ob>:<ih-'cococob5to H

en • » o

to ,_, . co to co to to to co to co to j-> M "£' >

h-> i co !*>. o bo rf>- If*. J-i ^ to •<» en ° n' t"

to ,_, *• to to to to to to to to to co j-i to
'-q to o to en «o «o co rf». to «o co J-" ^ M

5?

to . . co to co to to to to to co to to £. j^

^i iib^bbobjbobsenbjb10

os 2. o

to ;>;;K>COCOtOtOCOrfk.tOt3aQQ

bj to i:i:-4bb»b>rf^btOMw

to g o

to ,_, ...cototOMbotocotototoS. n

rik. i en en ^ «o bo o> en bi to *" g

co •»

to . . co to to to to to to to M 5'

co co ::::rf^£to>>ffctocnO)Cn St

^

to ....... to to to to to to ff

to t>

to co to to to _to to g.

b» en iiiJ^.'h-'enenco^iQ

oo «>

tO . . ; ; ; ; ; • JO tO tO g S,

en co iiiiiiliiitkOkfc n

o 8.

to to to to «o

bo co ; ; •;;:;;: bo bo bo

•

^ ••••;:::: .M ?» -10 o

co eo :::::: co >-> «o

448

MISCELLANEOUS STUDIES

was reached are probably within the experimental error. The large
difference between the total solids and sugar noted during the first
stages of ripening is no doubt due to the high acid content of the
unripe grapes. The fact that the difference remains fairly constant
after the grapes have become mature is to be expected, because the
cream of tartar, total acid, and protein remain fairly constant as
maturity is approached and during the periods of maturity and over-
ripeness.

• 6

I-

Van'af/oa /a

Thr

A/oft

os

/OOcc

go 3o •TO oo oo /o ov yo /oo /TO
Fig. 11 — Variation in non-coagulable protein content for three varieties, 1916.

Protein. — The total nitrogen content of the various samples was
multiplied by 6.25 to convert it into its protein equivalent. Owing
to the fact that the samples were sterilized by heat and filtered before
analysis, the figures represent only the protein not coagulated by heat.

The curves show that there is a slow increase in protein content
during growth and ripening and the greatest increase occurs during
the period of most rapid increase of sugar and most rapid decrease
of acid. The increase amounted to about .2 per cent in the case of the
Muscat and .6 per cent in the case of the Cornichon. The increase
seems to be quite definite, although the protein curves are not so
regular as those of total solids, sugar, and total acid.

[128]

CHEMICAL COMPOSITION OF GEAPES 449

SUMMARY OP CHANGES IN MUST OP GRAPES DURING GROWTH AND
RIPENING OF BERRIES

1.. Total Solids.— The total solids remain fairly constant during
the period of growth, corresponding to the period between setting of
the berries and the time at which the berries have reached almost
full size but are still hard and green. From this point on, there is a
rapid increase in total solids due to increase in sugar.

After the period usually considered as full maturity is reached,
the increase in total solids is slow. The question may be raised as to
whether this last increase is due to an actual synthesis and secretion
of sugar or other solids, or simply to evaporation of water. The fact
that there is no change in the curve of the acid decrease at this time
indicates that the same processes are continuing and that the increased
Balling degree represents an actual increase of solids. This view is
fortified by observations regarding the increase of weight of solids
during the ripening of raisin grapes. It has been shown that the
weight of dried grapes shows a continuous increase up to the highest
degree observed, 28.75 Balling.7

2. Sugar. — The total sugar during the growth period comprises
only a small amount of the total solids. During ripening, the sugar-
rapidly increases and then constitutes a much greater proportion.
During ripening, the sugar curve follows the total solids curve closely.
It is more or less the mirror image of the total acid curve multiplied
by five, i.e., increases as the acid decreases.

3. Total Acid and Free Acid. — During the early stages of the
growth of the berries, the acidity increases owing to an increase of
free acid. This is a fact that the authors have not found mentioned
in the literature. During ripening, the total and free acid rapidly
decrease. After maturity is reached, the decrease is very slow.

4. Cream of Tartar. — There is a very slow, but usually fairly defi-
nite, increase in cream of tartar during ripening. This increase is
very much less than the decrease in free acid, and therefore can not
account for any great part of this decrease.

7 Bioletti, Frederic T., Relation of the maturity of the grapes to the quantity and
quality of the raisins. Proc. Inter. Cong, of Viticulture, San Francisco, 1915,
pp. 307-314.

[1291

450 MISCELLANEOUS STUDIES

5. Protein. — The protein not coagulated by heat increased defi-
nitely during growth and ripening, although the increase was not so
regular nor so marked as the increase in sugar or the decrease in total
acid.

6. Difference Between Total Solids and Sugar. — This factor re-
mained constant for the lower percentages of total solids, decreased
during the rapid ripening stage, and remained constant through
maturity and over-ripeness.

[130]

X

Q

111

California. University

Miscellaneous studies in
agriculture and biology

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY