THE AMERICAN NATURALIST

THE

AMERICAN NATURALIST

A MONTHLY JOURNAL
DEVOTED TO THE ADVANCEMENT OF THE BIOLOGICAL SCIENCES

WITH SPECIAL REFERENCE TO THE FACTORS OF EVOLUTION

VOLUME LHI

NEW YORK
THE SCIENCE PRESS

1919

a
AA

n
f

THE
AMERICAN NATURALIST

Vot- DIIIL. January, 1919 No. 624

THE INHERITANCE OF HULL-LESSNESS IN
OAT HYBRIDS!

PROFESSOR H. H. LOVE anp G. P. McROSTIE

(IN COOPERATION WITH THE OFFICE OF CEREAL INVESTIGATIONS, U. 8S.
DEPARTMENT OF AGRICULTURE)

Tue purpose of this paper is to set forth some results
obtained in certain crosses between the hull-less types of
oats and some of the hulled forms.

The hull-less types belong to the species Avena nuda.
The origin of these forms is not definitely known, al-
though, according to Carleton,? they appear to have
come from central and eastern Asia. Reports are to the
effect that a certain form of this oat has been cultivated
in China for a thousand years or longer. These types
are not generally cultivated in any other countries and
perhaps due to this fact little study has been made of the
various hull-less types and their possible origin. From
some observations made on our material it seems quite
possible that these hull-less forms may have originated
through mutative changes. For example, in a pure line
of the variety Sixty Day certain spikelets suddenly ap-
peared which were very similar to the true hull-less
forms in having the caryopsis loosely held in the glumes
and an increased number of flowers per spikelet. These
seeds were tested, but did not reproduce this hull-less
tendency. A large number of similar cases have been ob-
served particularly with hybrids, although none of these

1 Paper No. 68, Department of Plant Breeding, Cornell University, Ithaca,

N: y.
2‘*The Small Grains,’’ 1916.

6 THE AMERICAN NATURALIST [Vou. LITI

has been tested as to its inheritance as yet, but it is
planned to do so.

Trabut® says in regard to the possible origin of hull-
less oats that ‘‘the study of the domestication of Avena
presents, from the genetic point of view, some rather
substantial arguments in favor of an ambient medium, a
modifying agent causing fluctuations which end in the
formation of varieties well characterized and fixed by
selection.’’ |

Previous Work

A number of investigators have studied hybrids be-
tween varieties of Avena nuda and Avena sativa. Nor-
ton‘ was the first investigator in America to hybridize
these forms. He reports that
the spikelet of the naked varieties usually has more than three grains,
while in the hulled types three grains is the limit. The first generation
plants produced a head naked at the top and hulled at the bottom. In
the second generation, one fourth of the progeny were typical naked
plants, one fourth were hulled, and one half like the first generation
hybrids. The naked plants all had long spikelets with more than three
grains, while the hulled plants had spikelets with the usual two or three
grains. In future generations no exception to this rule has been found
except that one second generation plant of a cross between European
Hull-less and Garton’s Tartar King which seems to have become fixed in
the intermediate hybrid type. In this example we have an extremely
rare case of the fixation of a heterozygote or hybrid type.

Gaines® reports having made some hybrids between
hulled and hull-less oats and first separated the F, types
into two groups, hulled and hull-less. In making such a
grouping the heterozygous types were put into the group
which it resembled most. He obtained from one cross
48.7 per cent. hulled and 51.3 per cent. hull-less plants
and says, ‘‘this indicates an intermediate about half-
way between hulled and hull-less for the heterozygous
types. As was mentioned above, the separation was
made arbitrarily into hulled and hull-less, according to
the type any given plant most nearly resembled.’’

3 Journal of Heredity, Vol. 5, p. 84, 1914. Translation of original paper.

4 American Breeders’ Association, Vol. III, p. 285, 1907.

5 Washington Agr. Expt. Sta. Bul. 135, p. 58, 1917.

No. 624] INHERITANCE OF HULL-LESSNESS 7

In another cross Gaines found 77.1 per cent. hulled to
22.9 per cent. hull-less, which caused him to conclude that
the two crosses were not similar in their behavior. This
would seem to be the case from the data at hand, yet in
1914 Gaines made a number of other crosses, among
which according to the pedigree numbers is another one
between these two sorts, Black (Wash. No. 665) and
Hulless (Wash. No. 680) which gave, this time, results
very similar to all the other crosses reported as made
that year, which indicated a 1:2:1 ratio. Gaines did not
offer any explanation as to the different behavior of these
two crosses between the same two sorts.

From these experiments Gaines concludes,
the percentage of hulled type suggests a simple Mendelian recessive
although in every case there are a few too many hulled plants. The
percentage of hull-less plants is not only very irregular in the different
erosses but is also irregular in the different families within the same
cross with the exception of the two families of Sixty Day X Hull-less,
which gave a ratio approaching 1:2:1. The intermediate types showed
great variation. Plants could be found with only one or two spikelets
that showed the hull-less character. Others could be found that showed
the hulled character in only one or two spikelets, and plants were ob-
tained with every degree of hull-lessness between these extremes. How-
ever, most of the intermediates produced more than half hulled oats.
A eurve fitted to these intermediate variations in Black Tartarian
X Hull-less shows larger numbers at either extreme and few numbers
showing per cents. of hulled oats ranging from 30 to 50. This is just the
opposite of what we would expect if the hull-less character was caused
by a single Mendelian unit which produced an intermediate in the F,.

In a paper by Zinn and Surface® results are given of
a cross between a hull-less and hulled oat. The sorts
used were Avena sativa patula var. Victor, and Avena
sativa nuda var. inermis. The results indicate that their
forms agree very closely with those reported by Norton
and Gaines. The following paragraph gives part of
their conclusions.

The F, generation is distinctly intermediate in most characters. In
regard to the glumes, both naked and firmly hulled grain as well as in-
termediate forms are found on the same panicle and even in the same

6 Journal of Agricultural Research, Vol. X, No. 6, pp. 310-311, 1917.

8 THE AMERICAN NATURALIST [Vov. LIII

spikelet. As shown in Table I, the spikelets near the top of the panicle
are either entirely naked or nearly so, while those spikelets near the
base of the panicle tend to be firmly hulled. A similar but less marked
relation is to be observed between the spikelets at the tip and base of
each whorl.

In the F, generation a large number of intermediate forms appear.
In addition to the two parental hull types, four intermediate classes
were distinguished. These intermediate forms contain all gradations
from the plants with perfectly hulled grain to the perfectly naked
forms.

As shown in Table II, the inheritance of the hull characters presents
a simple Mendelian relation giving 1 hulled, 2 intermediate, 1 naked.
Likewise, in respect to grain color, there are 3 black plants to 1 white,
in the second generation.

MATERIAL AND METHODS

In connection with some experiments in oat breeding
a number of hybrids between the hulled and hull-less
forms have been made. While in these crosses the in-
heritance of other characters such as color of glumes,
pubescence, awns and the like, are very interesting, the
present paper will be confined to the discussion of the
inheritance of the hull-less and hulled characteristics.
A more complete discussion of the various characters is
being prepared for a later publication.

e authors want to take this opportunity to express
their appreciation for the valuable assistance in note-
taking and tabulation of results rendered by W. T. Craig
and Miss A. M. Atwater. Their work has been of great
aid in conducting these experiments.

The hull-less oat used for the various hybrids was
typical of the Avena nuda group and differs from the
Avena sativa forms by three important characters: (1)
The lemma, or flowering glume, and palet do not clasp
the kernel as in other forms, and the kernel is therefore
loose, or free, within the hull; (2) The rachille of the
three to many-grained spikelet are so elongated that
the uppermost grains are borne above the empty glumes;
(3) The glumes and the lemmas are similar in texture.
The illustration (Fig. 1) will give a fair idea of the par-

No. 624] INHERITANCE OF HULI-LESSNESS

\ fare hull-lege sprkelet
Fic. 1. THE FEMALE PARENT (Avena nuda) USED IN SERIES 382.
ticular characteristics which separate Avena nuda from
the hulled species. This type is typical of the hull-less
forms used in the following hybrids.

The hulled forms were all varieties of Avena sativa
with one exception. In this case Avena fatua was used

À
l Ñ
A Vy he
f
pone”:
AMANN

Fig. 2. THE MALE PARENT (Avena sativa) USED IN SERIES 382.

as the hulled parent. Fig. 2 shows the hulled variety,
Sixty Day, used in one of the crosses. It is typical of a

hulled oat.

The first crosses of this sort were made in 1910 and
the F, plants grown in the greenhouse the following win-
ter. The first cross was between Hulless and Black Tar-

No. 624] INHERITANCE OF HULL-LESSNESS 11

tarian. Here the hull-less form was used as the female
parent. The male parent possessed, as indicated by its
name, black glumes, and was a typical hulled oat. The
other cross was between Danish Island and Hulless in
which the hulled type was used as the female parent. In
each case the F, type was typical of the F, types as de-
scribed by the authors mentioned above. This form is
intermediate in that both kinds of kernels, hulled and
hull-less, are found on the same head. The type of pan-
icle resembles the hull-less parent more than it does the
hulled and may be considered as intermediate in type.
There are some spikelets with hulled and some with hull-
less kernels and also some with both hulled and hull-less
kernels. As a usual thing the hulled spikelets occur
towards the base of the panicle while the hull-less kernels
occur near the terminal spikelet which is almost inva-
riably hull-less if such kernels are present at all in the
panicle.

There are fewer hulled than hull-less kernels on the
F, types. The percentage of hulled kernels does not usu-
ally run very high. In Fig. 3 is shown a typical form of
F, panicle of a cross between a hulled and hull-less oat.

The F, generation of these two crosses were grown in
the field in the summer of 1913. The plants were then
sorted into two groups, hulled and hull-less, or hull-less
like. All those plants having any indication of hull-less-
ness were placed in the hull-less class. The result of
these counts was as follows:

Bocce a Varieties Crossed Hull-less | Hulled
111 Hulless X Black Eg aig Si Se rer ne. 129 37
51 Danish Island X Hulless. .:. 2... eo cs: 364 93
- 493 130

Although the ratio deviates considerably from 3:1 it
indicates that this character behaves as a simple mono-
hybrid and that there is one factor pair concerned. In

12 THE AMERICAN NATURALIST [ Vou. LIII

Fie. 3. F, INTERMEDIATE TYPE OF A CROSS BETWEEN HULL-LESS AND HULLED.
Series 382,

order to test this out more fully all plants from which
good seed could be obtained were grown in the following
year. The results obtained from these plants showed
without doubt that the segregation followed a simple
monohybrid ratio. That is, the pure hulled and hull-less
plants bred true to these characteristics, while the inter-

No. 624] INHERITANCE OF HULL-LESSNESS 13

mediate types reproduced the three types again. The
second generation plants tested as to their composition
gave the following results in the third generation:

Seri Pure . Inter-
rl Varieties Crossed Gelies |. mediat, | Hulkias
lil Giuliess X Black Tartarian Iii. SEn 37 85 38
51 Danish island X Hualless -norak 115 216 114
152 301 152

It is apparent from these results that certain plants were
classed as intermediate in the second generation, which
were in reality pure hulled plants. It is evident that the
hulled-hull-less character is inherited in a simple Men-
delian fashion so far as its general behavior is con-
sidered.

The heterozygous plants produced in the third gen-
eration were examined as to the relative amounts of
hulled and hull-less kernels present. This was done by
threshing a representative head from each plant by hand
and counting the hulled and naked kernels and express-
ing the result as the percentage of hulled kernels.

In order to determine whether the results from a
single head fairly represented the type of the plant a
number of plants were examined and recorded a head at
a time. The percentage of hulled kernels for a represen-
tative head was then compared with that for the entire
plant. The average percentage was the same for the
results from single heads as it was for the entire plant.
Although there was some deviation in the individual de-
terminations, the correlation between the two methods is
very high.

The result of determining the percentage of hulled
kernels was to indicate the great variation existing,
which was from a very low to a very high percentage.
As a result of these observations it was apparent that
while in the hybrids under consideration the usual 1:2:1
ratio was observed, some factor or factors were pres-

14 THE AMERICAN NATURALIST [Vou. LII

ent which affected the heterozygous forms in such a way
as to modify the amount of hulled or hull-less kernels
present.

In order to determine this effect in a more definite way
it was planned to sow seed from heterozygous individuals .
which differed as to the percentage of hulled kernels
present. A rather large number of such seeds were
planted in 1915. The plants were severely injured by a
storm, so that accurate pereentage determinations could
not be made.

In the meantime, however, a number of other crosses
had been made in which the Avena nuda was used as one
parent. The following sorts were crossed with the naked
oats: Swedish Select, Sixty Day and Avena fatua. Other
crosses are being studied but these will be reported on
later.

As regards the hull-less character the F, individuals
of these crosses were all similar to the description of the
first generation given earlier in this paper. Regarding
the other characters, the cross between the hull-less form
and Avena fatua showed some very interesting varia-
tions. These will not now be discussed.

Seeds of these various F, plants were sown and the
resulting plants harvested. From each plant a head was
saved and threshed separately by hand and the plants
then sorted into hulled, intermediate and hull-less. The
result of the several crosses is given here:

Series Inter- | Hull-
wa. Varieties Crossed Watled metae) tens
379 | Hulless X Avena fatua...... Mee bebe E ses 68 111 78
202° | Einlleds X Bwediush Select’... ffs bt Al 90 36
382 | Hollós X Bitty Dav. 6 ieee 75 193 53
Observed ois as A akara 184 394 | 168

Expotod aaae a 186.5! 373 186.5

The probable error is +7.98 and the observed num-
bers agree fairly well with the expected numbers. The
number of hull-less plants is too low and the number of
intermediates too high. It is possible that in some cases

No. 624] INHERITANCE OF HULL-LESSNESS 15

oR

Fic. 4. HETEROZYGOUS TYPE F, POSSESSING 10 Per CENT. OF HULLED KERNELS.
Series 382.

hull-less plants may have been recorded as intermediates
although the error from this source is not large. When
the results are considered on a 1:3 basis and the hull-less
and intermediates are grouped together we find that
there is a percentage of 24.66 +1.07 hulled plants.

16 THE AMERICAN NATURALIST (Vou. LIL

aa

Fic. 5. HETEROZYGOUS TYPE Fy POSSESSING 87.9 PER CENT. OF HULLED KERNELS.
Seri 82.

The results of these different hybrids show that hull-
lessness is inherited in a simple monohybrid manner and
that without doubt the difference between hulled and
hull-less oats in this regard is represented by one pair

No. 624] INHERITANCE OF HULL-LESSNESS 17

of factors. An analysis of the different heterozygous or
intermediate individuals of these second generation
plants showed that for these hybrids also there was a
great amount of variation in the percentage of hulled or
hull-less kernels in the individual plants, the variation
ranging all the way from less than 5 per cent. to 95 per
cent. or more. In Figs. 4 and 5 are shown two forms of
heterozygous plants, one very low and one very high, in
percentage of hulled kernels. These percentages were
obtained from the heterozygous individuals by sorting
the kernels from one head of each plant into hulled and
hull-less as outlined earlier. The percentage of hulled
kernels on the heterozygous plants of the second genera-
tion for the three series is given in Table I.

TABLE I

SHOWING PERCENTAGE OF HULLED KERNELS ON THE HETEROZYGOUS PLANTS
OF THE SECOND GENERATION IN CROSSES BETWEEN HULLED
AND HULL-LESS OATS.

Percentage of Hulled

Number! Varieties Crossed 19! 19] alala al 19/99! alag 9! 9] 8} 2} 9
> 7“ AN BRIAR AY AY Ae! Al Ye
OO Hi Hw) oO) Oo] ©

72.5
7
2
7.
2
97.5

379 |Hulless X Avena
Nabe ak 2| 8) 5| 4| 4) 3| 4| 4| 7| 5) 5) 8) 4

202 |Hulless X Swed-
ish Select ......|10} 4| 4| 3| 5| 8| 2| 1| 7| 6| 3) 4) 2| 5
382 |Hulless X Sixty
Bee ee ces

o o
ao
~J
—
©
p
w
—
or

~J
A
g=
pt

5 5 5) 4) 4| 6 6| 820/152221 (15/13/1711

15/1112 17/21/22

From this table it is seen that there is considerable dif-
ference in the percentage of hulled kernels on the dif-
ferent heterozygous plant. The range is from a very
low percentage or one which indicates nearly all hull-less
to a very high percentage or one which is nearly all
hulled. There is no general grouping near the middle of
the series, as might be expected with the exception of
series 382. This may be due to lack of numbers or to a
segregation of the different types which give percentages
ranging from low to high without any tendency to group-
ing. That it is not due to lack of numbers is probably

18 THE AMERICAN NATURALIST [Vor. LIII

borne out by the fact that in series 51, where over 900
plants of the third generation were sorted into the dif-
ferent classes, there was no indication of a grouping near
the middle classes, in fact, the slight indication of group-
ing was near the lower values. The distribution is as
follows:
TABLE II
SHOWING PERCENTAGE OF HULLED KERNELS IN THE HETEROZYGOUS PLANTS
OF THE THIRD GENERATION IN A CROSS BETWEEN DANISH
ISLAND AND A HULL-LESS OAT.

Percentage of Hulled Kernels Frequency Percentage of Hulled Kernels Frequency

0— 4.9 89 50.0 —54.9 42
5.0 — 9.9 80 55.0 —59.9 48
10.0 — 14.9 60 60.0 — 64.9 43
15.0 —19.9 49 65.0 —69.9 26
20.0 —24.9 59 70.0 —74.9 35
25.0 —29.9 48 75.0 —79.9 33
30.0 —34.9 47 - 32
35.0 —39.9 53 85.0 —89.9 26
40.0 —44.9 52 90.0 —94.9 21
45.0 —49.9 46 95.0 —99.9 15

It was planned to carry some of this work further to
answer in general two questions which are: (1) Does the
percentage of hulled plants obtained from any hetero-
zygous parent vary with the percentage of hulled kernels
possessed by that parent? (2) Do the hulled and hull-
less kernels of a heterozygous plant give approximately
the same results in their offspring?

In order to obtain data on these questions two of the
series have been continued. The hull-less-Avena fatua
series has not been carried further as yet but it is
planned to do so.

RESULTS FROM Serres 202—Swepiso SELECT X HULLESS

The first series to be discussed is the Swedish Select-
Hull-less cross. Seed from two hulled and two hull-less
plants of thẹ second generation were grown in the third
generation and each bred true to type. In addition to
these plants twenty heterozygous plants were selected
for planting. These varied as to the amount of hulled

No. 624] INHERITANCE OF HULL-LESSNESS 19

kernels. The range was from 3.2 per cent. to 92.0 per
cent. The number of seed was not large, therefore the
number of plants obtained was not as large as desired,
yet from the consistency of the results certain conclu-
sions are justified. The offspring from these twenty
plants were sorted into the three classes, hulled, inter-
mediate and hull-less. The intermediate plants were
again threshed and the percentage of hulled kernels de-
termined.

In Table ITI is given the percentage of hulled condition
in the parent plant, the segregation into the three groups,
the percentage of hulled kernels in the heterozygous off-
spring, the grouping into hulled and hull-less and (where
both hull-less and intermediate plants are grouped to-
gether) the percentage of hulled plants with the prob-
able error.

TABLE III
SHOWING SEGREGATION IN F, OF CERTAIN F, PLANTS TOGETHER WITH THE
CENTAGE OF HULLED SEED IN PARENT TYPE AND THE AVERAGE
PERCENTAGE IN THE HETEROZYGOUS OFFSPRING.

Segregation Obtained from Plants Sown and Resulting Percentage
of Hulled Kernels on Intermediate Forms.

ste of i a Bire Hull-

Heals | Hultet | atisto] aa | a Tatar: | tla | naes P" O te

in Plants mediate mediate

Offspring
202al-4....| 44.9 27 22 44.8 27 24.11 + 2
34 51 40 21.0 34 91 27.20 + 2.61
20.9 12.8 27 22.13
10 42 5 42 97 30.22 + 2.48
a. 92.0 42 62 17 52.6 42 79 34. x
22...) 65.1 13 36 7 60.6 13 43 23.21 + 3.90
25 90. 14 36 9 55.3 14 45 23.73 + 3.80
26 76.6 14 37 21 56.3 14 58 19.44 + 3.44
29 30. 12 19 15 44.1 12 34 26.09 + 4.31
31 56. 16 43 21 47.6 16 64
38 10.5 21 40 22, 21 66 24.14 + 3.13
40...| 82.4 19 33 19 53.6 19 52 26.76 + 3.47
46 16.7 14 24.7 14 42
51 56.6 41 10 47.1 25 51
60 74.3 T 10 4 46.0 7 14 33.33 + 6.37
s 3.2 10 26 32 22.7 10 14.71 + 3.54
TT 65.4 15 21 14 34.9 15 30.00 + 4.13
92 35.3 12 10 8 36.0 12 18 40.00
120 44.7 14 18 8 33.9 14 26 35.00 + 4.62
121 9.5 13 11 23 20.2 13 34 27.66 + 4.26
TOS.. 391 661 426 391 |1087 26.45 + .76
Expected. . 5| 739 | 369.5

20 THE AMERICAN NATURALIST [Vor. LIII

The results of the segregation into the three classes
gave 391 hulled, 661 intermediate and 426 hull-less. Here
the hull-less plants are too great in number, while the
number in the intermediate class is too small. The ex-
pected numbers are 369.5:739:369.5, with a probable
error of +11.23. It is possible that some intermediate
plants were classed as hull-less. Such a condition is pos-
sible since some intermediates are found bearing only
one or two hulled kernels, and if these should be lost
through shattering, such plants would be classed as hull-
less when in reality they are intermediates. When the
grouping is made into the two groups, hulled and hull-
less, it is seen that the 3 to 1 ratio is approximated very
closely, as there are 391 hulled plants to 1087 hull-less,
giving a percentage of 26.45 + .76 hulled.

An examination of this table shows further that some
of the families do not give ratios close to 1:2:1. This is
true with regard to certain families particularly with
certain of those coming from plants low in percentage
of hulled, and some of those relatively high in this re-
spect. The results of some of these families have been
brought together in Table IV.

TABLE IV

SHOWING SEGREGATION OF OFFSPRING COMING FROM SOME INDIVIDUALS LOW
OR HIGH IN THE PERCENTAGE OF HULLED KERNELS.

Percentage of Hulled Segregation of Offspring Into Different Types.
Family Num- Seed in Plants SHORTEN

a Sown Hulled Intermediate Hull-less
6 25. 34 51 40
9 20.9 27 35 60
38 10.5 21 26 40
46 16.7 14 20 22
66 3.2 10 26 32
121 9.5 B 11 23
T i 119 169 217
10 83. 42 63 34
11 92 '42 62 17
25 14 36 9
26 76.6 14 ot 21
40 82.4 19 33 19
Tow se: 131 231 100

No. 624] INHERITANCE OF HULL-LESSNESS 21

In this table the plants from parents having 25 or less
per cent. of hulled kernels give 119 hulled:169 inter-
mediate: 217 hull-less. From this result it appears that
those plants having a low percentage of hulled or high
percentage of hull-less kernels tend to produce a rela-
tively high number of hull-less plants. On the other
hand, those plants having more than 75 per cent. of
hulled kernels do not give results so striking. There are
more hulled than hull-less plants, yet not strikingly so,
and the hulled do not run higher than the intermediates.
It may be, however, that the degree of hull-lessness as
expressed by the percentage may influence the segrega-
tion in the following generations. This can not be def-
initely stated from this cross, and further evidence will
be needed.

TABLE V

SHOWING THE RELATION BETWEEN THE PERCENTAGE OF HULLED KERNELS ON
HE HETEROZYGOUS PARENT PLANTS AND THE PERCENTAGE
OF HULLED KERNELS ON ITS OFFSPRING

Percentage of Hulled Oats on Offspring
Percentage of Hulled a 2 alealale alelalelealelelelalalaiale
Oats on Plants @| a| $ Si í 8] F) S| 3] S$] 31S) 3] Si Fe) Si si 3 S
2/4) s) 3) 33,2222) 3)s) 52 223 )3
“S| Si 8| 8| S| 3| S| 3| S| $| S| Sie] £ 8] Si 8} 3
2 12, 3} | 3/1 H i 111) | 1/1) | 26
5 1| 1| 2 2) 2 1} 1
10.5 5| 2; 3i 6| 6| 1,1) 1) | 1 26
16.7 A3 | 3i 4) 21/2) | 2 1 20
20.9 910| 7| 4 1} 1 35
25.0 5112| 4! 6| 5| 2| 9| 2| 4| 1 1
30.7 1} 1} | 1 2| 2| 2| 2) 1| 1| 1| 1| 21 )
35.3 a} inuiti 1 1 )
44.7 UHI 2 4| 3 3} 2) ] }
44.9 4' 31 6, |3' 1| 6| 3| 2| 3| 4! 7| 7! 6| o| 6| 2 63
56.6 1} 3i t 3] 3, 2} 6 3} 3} 2} 3ks) 1) al ala) fal
56.8 5| 1| 1; 4| 3| 2| 2| 4| 2| 5| 6| 2| | 3|- 2| 1 43
65.1 2 3 6| 4| 2| 4 2| 6| 4) 1| 2| |36
4 1| | 1} 1} 1) 4 2 1 1 1} | 21
3 1} 1} | 1 1 1} 1} | 1/1 )
76.6 1] 2! 3} 2} 1) 1) 3] 81 41 5} | al ale 1| 37
32.4 1| 1| 2| | 2| 3| 3| 4) 3| 3| 1| 1| 2| 2| 2| 1| 2 3
33.0 5| 1| 2| 8| 4| 5|. 2| 7| 5| 2| 5| 3| 3| 5| 2| 1 1 3
90.3 4 2) 2 1 3} | 31 1] 7| 11 5] 3] 3} 5
92.0 1| 4| 2| 1| 3) 3| 2| 3| 2| 5 5} 6| 4| 2| 4| 4| 7| 1| 3| | 62
50 60 37:39 41'39/39130 32136 39 42 32'34'18 32'32.15112] 2 661

22 THE AMERICAN NATURALIST [Vou. LIII

The relation between the percentage of hulled kernels
on the parent and the percentage of hulled kernels on the
heterozygous offspring for this series is shown by the
correlation table given above. The correlation coefficient
is. 421 + .022, which shows a very definite relation be-
tween the percentage of hulled in the parent and offspring.

This relation is also better shown by means of a curve
(Fig. 6) in which the parents are represented by the
dotted line beginning with the lowest and increasing to
the highest value. On the same ordinate is plotted the
average value for the heterozygous offspring, and to this
line has been fitted a straight line whose equation is
y = 20.1999 + 1.95792.

Certain of these families show a decided grouping; for
example that represented by 20.9 per cent. shows a de-
cided tendency to be grouped in the lower classes, while
that represented by 65.1 per cent., with five exceptions,
shows a grouping around the higher classes.

Resvuuts oF Serres 382—Srixty Day X HULL-LESS

From the second generation of this cross six hetero-
zygous plants were selected for further study. The re-
- sults of three of these will be discussed here. These
plants possessed different amounts of hulled kernels,
which expressed in percentages were as follows, 73.3,
37.7, 49.3. Thus, there was one high, one low and one
medium plant. The offspring of these gave the follow-
ing results when grouped in the three classes:

Hulled Intermediate Hull-less
A E 23 ! 55 20
t 15 37 24
ao oo 53 105 49
91 197 93

These figures agree very well with the expected 1:2:1
ratio. Single heads of the heterozygous plants of these
three families were threshed and the percentage of hulled

No. 624] INHERITANCE OF HULL-LESSNESS 23

kernels per plant determined as before. The three fam-
iles gave the following distribution:

Eta
Bek
Pore a
Bee
S
Pye
Er |
ER
Hi

E
E
ad
E
+

eh eee!
grain on the heterozygous plant used as parent and the average percentage of
hulled kernels on its heterozygous offspring. Dotted line represents the value
for the parents and the solid line that of the heterozygous offspring. Series 202

From these distributions it is clear that the percentage
of hulled kernels on the parent form influences the
amount of the hulled condition. The average percentage
of the offspring in each case agrees closely with that of
the parent forms. |

From these three families several plants differing in
their percentage values were selected to continue the
study in the fourth generation. As observed from the
frequency distribution just given it is noted that series
7 is of high value, while 8 is relatively low and 9 varies
from very low to high. The plants selected then in gen-

24

THE AMERICAN NATURALIST

[Vou. LIH

eral represented the types of their lines; that is, those
from 7 were generally high, those from 8 generally low,
and those from 9 both low and high. The offspring of
these various selections are arranged in a table similar
to that for series ‘202.

Per- Percentage of Hulled in Offspring aa Per
cent- Cent.
: ag z of
Series No. Hulea) | alol elal elelelaleļe|ele|e| e| e| a| e| 2| 2| $ | nuna
in| 2| o| $| S| X| ŠIS S| $| S/S] S| S| 2| S| 2| S| 2| S| 8| & | Ker-
Pianta | $| $| 2! 2A eji 2] 3) 2/4) 4) 3) A) 3) 4) 3) 2) a) a) [nes on
ajo) sel sisi gig! S| S| 3/8! $| 3| 5| £| £ 3/3/83] | apring
382al-7 | 73.3 1 1| 3/6 5| 5 6/12/10, 6) 1 55| 67.3
8} 37.7 |1|1;2)3]| 3; 641512; 5) 2) 1) 2 37| 33.7
9| 49.3 |1\/1/3/6}| 7 6|5|4| 4/14) 5| 811/11 7| 8| 1] 1] 2 105) 50.1
22 5 911129 / 9/7 |2212/14/18 17/1918| 7| 2| 2
TABLE VI

SHOWING SEGREGATION IN F, OF CERTAIN F, PLANTS TOGETHER WITH THE
PERCENTAGE OF HULLED SEED IN PARENT TYPE AND THE AVERAGE

PERCENTAGE IN THE HETEROZYGOUS OFFSPRING

Segregation Obtained from Plants Sown and Resulting Percentage
of Hulled Kernels on Intermediate Forms.

Per Per

ulied Inter- | Hull Balled jase ana: | Por Genk: thutiea

Kernels| Hulled} mediate} less Inter-| Hulled| Inter- and P. E.

in Plan mediate mediate
So Offspring

$6241-7-10. .| 63.6) 22 25 63.1 22 56 28.21 + 3.3]
11.. 550 | 24 t 10 46.0 24 44 35.29 + 3.54
$2: 4 71.0 | 28 3 22 70.5 28 60 31.82 + 3.11
.33..| 70.3 21 3 26 62. 21 79 21.00 + 2.92
37.. 500 8 19 57.5 8 50 13.79. + 3.84

AS, | 817|- 18 7 13 75.1 13 40 24.53 + 4.(
55..| 68.0} 19 ) 28 67.8 19 67 22.09 + 3.15
8-14..] 43.6 | 26 L 19 37.6 26 53 32.91 + 3.29
17| 392| 28 3 23 4 28 86 24.56 + 2.74
Py Zee 16 } 14 25.3 15 50 | .23.08 + 3.62
Pele WE: 4 ) 9 19.2 4 18 18.18 + 6.23
Po 008] 12 30 13 49.8 12 43 21.82 + 3.94
35: | 179] 22 43 24 20.5 67 24.72 + 3.10
61..| 59.6 | 13 11 50.6 13 38 25.49 + 4.09
64. 9.8] 16 j 14 25.8 16 50 24.24 + 3.60
9- 9..; 14.0 4 ) 7 19.2 4 36 10.00 + 4.62
44 62.4 | 24 ) 15 49.1 24 65 26.97 + 3.10
04) ol 15 | 20 22.2 15 4l 26.79 + 3.90
166.. 797| 16 3 22 1.2 16 49 24.62 + 3.62
176. .| 879: 19 32 17 67.2 19 49 27.94 + 3.54
Tots. oc: 349 | 690 |351 349 1041 25.11 + .78

Expected | 847.5| 695 | 347.5 347.5 |1042.5

No. 624] INHERITANCE OF HULL-LESSNESS 25

The total number of plants in the hulled, intermediate
and hull-less classes agree very closely with the expected
numbers. The same is true for the 3:1 grouping, since
the percentage of hulled is 25.11 + .78, which shows with-
out doubt that the various families give offspring which
follow the 1:2:1 expectancy.

In this series there is little evidence that the percent-
age condition of the parent plant affects the type of seg-
regation in the following generation. In sini the
segregation of the various families follows a 1:2:1 ratio
regardless of the percentage condition of the parent.

To show the relation between the hulled condition of
the parent forms and that of the heterozygous offspring
a correlation table was made in which the different fam-
ilies were arranged according to their percentage values.
The coefficient of correlation here is .726 + .012, which

TABLE VII
SHOWING THE RELATION BETWEEN THE PERCENTAGE OF HULLED KERNELS ON
THE peen tex PARENT PLANTS AND THE PERCENTAGE OF
ED KERNELS ON ITS OFFSPRING

Percent- Percentage of Hulled Oats on Offspring
Hulled elieiglaigigiziziaigigizizizieiaie|2
oatson|e|}S/2/2/8i8l/sieleisigisi¢isizielsigig¢is
me S| Alaa aliaa Eaa
PIS/SI8(R/8/8/S/3/S/8/S/8/F/2/28/18/8)/8
9.8 | 2| 2| 4| 5| 4] 5| 7| 2| 3] 2 36
14.0 | 2| 7| 7| 3| 1| 2| 2| 1| 2| 2 29
66 a 9
17.2 | AE 4| 5| 3| 1| 3 43
23.2 | 2| 2| 4| 5| 4| 7| 2| al 2 36
23.6 | 2 6j. 7| 2| 1) 1 1 20
38.2 2l 3/12 11| 11] 8| 7! 3| 1! 3 2| 63
43.6 2| | 5| 2| aļ 5| el e| 1| 2| a - 34
53.0 1) 1] 2} 1] 3} | 2] 2] 6 5} 63) 2 a) fa 34
59.6 1) | 4! ol 4! gi 4) ol al 4 27
60.8 1 2) 1) 3} 4) 3}: 6 2) 3) 3) 2 30
62.4 | 4| 2| 2| 1 1} 4| 2| 3] 5] 9| 6 4| 4] 1 2| | 50
63.0 | 1 2 1) 2} 5| 1} 6| 5| si | 1] af | 30
63.6 1 2} 1} 1} 2} 2| 2} 1] 9} 1] 3] 6 1 31
70.3 i] 1] 2} 2 | 7} 9] 7 Of S| 6 1 3 53
77.0 1 1 1) | 4| 1) 5] 9} 6 5] 3} a} | 37
79.7 Hini 2| 1| 4| al 1/10| 2| 27
80.0 | 1| 1 3 1} 3} 4 2) 3] 6 2} 3) a) a) fa) an
81.7 a) dat al bh al a) ai- 27
87.9 1 EE EE sl al a 31
17| 22| 29| 37) 31| 35/45) 41| 42| 46| 54| 49| 34| 57| 38| 40| 31| 17| 17| 5 | 687

26 THE AMERICAN NATURALIST (Vou. LIII

is considerably higher than it was with the 202 series.
One reason for this may be that perhaps there is a dif-
ference between this series and the former or that the
result is caused by grouping the three families. When
a correlation table is made for each of the three families
ER se oe ee Se eee
HS oe (ee ee
SERS

EEL
:

Hg Eg REH
Senate ea

i
i

n E E A D
IHE HES egy

KEN M

HERS A ER RET RHH

jii

j

A

ii
HEHA RE
an
i

GE EEH a EA iH

8 i ETTAN

HS ET EE
À HAAN MAHR
{ies UM Tes RP

NY

FAR ect A
HEREN

HMM Ge ea
AN
euhhh

Me
EEN i

i EN

if
Hi!
ANCE
Nt
i
t

BH l:
iliii

fiil
eli

i Uma GE E E

A SN

i Ra

i
[i

SCHEER Mea l a AH REAA

UG EF i e TRH

E
pa
peg
E
a
‘ E E

ilk
ietan aT A
À ee |

fat

EET Ha A

RH
i
E
i a
en

Ba a
a a

ANGH SIHI NIRA HHIH IAA RT

EE E E e EN

aie
A A DE RNE RN

ith
tHE

:
X

EEE He FR REN Fi;
E E E

a TR
el

Hab’

H
p

i
EES a SG E a

i
a gg
Tn ae

4

i]

mi
e oereerea i
REEE EE] iba

it
Hil

Be Bal

i Aa

di
Bane
nied

FREY FD i EEH RA a Eh Alm aM E THEE EH

HTD FASTEST ET H CEER EN RT

f f}
"E BRAR
IA
Hi
£
H

H

ded HEH ib UEU LAH Pe ER

AHO GNERRA INN ER EHTE

ae E N
AETV RUA RHH

RR i=

=
E=]
=
am: ES

IG. 7. These curves show the relation between the percentage of hulled
grain on the heterozygous plant used as parent and the average percentage of
hulled kernels on its heterozygous offspring. Dotted line represents the value
for the parents and the solid line that of the hetrozygous offspring. Series 382.

separately, correlation coefficients of .296 + .039, .623
+ .025 and .741 + .024 are obtained. Thus it is seen that
within any family correlation exists to a greater or less
degree. Putting the three families in one table does in-
crease the correlation somewhat over the average value
for each alone. Another and possibly more plausible
reason is the fact that these plants are of the fourth
generation, while those of 202 are of the third.

No. 624] INHERITANCE OF HULL-LESSNESS 27

The three parent series from which these were taken
were of three types as mentioned before, therefore the
parent plants selected from them carried the tendency to
produce high or low as the case may be, and when they
are all arranged in a correlation table naturally a high
coefficient is obtained. In other words, the three parent
forms were more nearly homozygous, so to speak, for
high or low values. More will be said on this point later.

This relationship was further shown by means of a
graph showing the relation between the parent percent-
age condition and the average value for the offspring,
the same as was done in Fig. 7.

In this case the relationship is higher than in the
former series. The equatioa to the straight line is
y = 17.2411 + 3.20622.

That the plants arising from heterozygous plants hav-
ing a low or high percentage of hulled kernels did not
segregate in a manner indicating any influence of the
hulled condition of the parent plant, as was the case to
some extent with series 202, is shown in Table VIII.

TABLE VIII i
SEGREGATION OBTAINED WHEN SOWING SEED FROM HETEROZYGOUS PLANTS
Havine Low oR HIGH PERCENTAGES OF HULLED SEED

Percentage of Hulled Kernels Segregation of Offspring Into Different Types
in Plants Sown ra ORINA eee

sp 16 34 14

14.0 4 29 -

17.0 4 r :

23.2 15 14
23.6 15 21

76 f2 88

77.0 28 $3 er

79.7 16 27 22

sèn - 32 17

84 | 155 93

It is clear that the percentage of hulled seeds does not
seem to influence the segregation as far as these data are
concerned.

28 THE AMERICAN NATURALIST (Vor. LII

HULLED COMPARED WITH HULL-LESS KERNELS

In order to learn whether there was any difference in
the ratios produced by the hull-less kernels from the
heterozygous plants the seed from the third generation
plants used was separated and planted separately; that
is, the hulled and hull-less from family 382a1-7—45 was
planted separately so that the ratio may be determined
on each lot of plants. This was done for all the families.
These results are given in Table

TABLE IX

RESULTS OBTAINED FROM SOWING HULLED AND HULL-LESS iara FROM THE
SAME HETEROZYGOUS INDIVIDUAL SEPARATEL

esis Obtained from Segregation Obtained from
Hulled Seeds Hull-less Seeds
Family No.
Hulled | Intermedi-| Hull-less Hulled Intermedi-| Hull-less

Plants ate Plants| Plants Plants (ate Plants Plants

382al-7- 45...... 11 26 12 2 1 1
fog | E 18 22 20 4 9 5

F eo apne eae 19 27 10 5 t 0
Mas ates 5 18 28 az 10 10 5

Sess ae ee 20 39 21 1 14 5

7- 37 ae 8 25 19 0 6 0
FB 6 en% 14 31 25 5 8 3
STA 18 19 8 8 15 11

Be Hra 10 28 12 18 35 11
2a O 3 ra 4 12 29 10

8- 23.. ki 6 6 0 3 3

S- 26, ee 9 18 8 3 12 5

8+ BO, ceca 5 16 7 17 27 17

8- 61. 9 19 9 4 8 2

8- 64 3 7 + 13 29 10

Lede 3 6 3 1 23 4

9- 44...... 16 35 13 8 15 2

a ote 7 8 7 8 13 13
9-166...... 13 22 14 3 5 8
9-175 19 30 16 0 2 I
ete cee 227 419 235 122 | 271 116

; Ehana ah i AER aS 220.25! 440.5 220.25| 127.25 — 254.5 127.25

In many of the cases the numbers are too small to give
good ratios, yet the important point is obtained from
the summation of the two series. In each case these agree
very closely with the expected numbers. If there was
any difference we might expect the hulled kernels to pro-
duce relatively more hulled plants and the hull-less rela-

No. 624] INHERITANCE OF HULL-LESSNESS 29

tively more hull-less. The facts are the reverse. In the
series from the hulled kernels the hull-less plants are
in the majority and the opposite is true for the hulled
plants from the hull-less seed. It is very evident that
there is no relation between the kind of kernel (hulled
or hull-less) sown from a heterozygous plant and the off-
spring produced.

GENERAL DISCUSSION

From the foregoing data it seems without doubt that
the inheritance of the hulled condition follows a simple
Mendelian ratio giving in general 1 hulled, 2 inter-
mediate, 1 hull-less. This is in accord with the results
obtained by Norton, Gaines, Zinn and Surface, and
others.

In regard to the relation between the hulled condition
of the heterozygous parent plants and of the offspring,
it is clear that there is a very close agreement in regard
to the hulled percentage. When high or low plants are
selected they produce heterozygous offspring giving high
or low percentage. In most cases, however, the usual
1:2:1 ratio is obtained. This is true in general in all
cases of the 382 series but not so for 202. Whether the
202 series behaves differently or whether in reality it
will agree with 382 will have to be determined with
further work.

The percentage relation shows that there is a variation
from very low to a very high percentage. Owing to this
fact and that any heterozygous plant tends to reproduce
a simple monohybrid ratio, in which the heterozygous
plants tend to follow the percentage relation, it seems at
first that we are dealing with a case of multiple factors,
in which one primary factor pair determines the hulled
or hull-less condition and the other factors influence the
hulled condition of those plants only that are hetero-
zygous for the primary factors. This may be so, as the
results of selecting high or low individuals seem to indi-
cate. If, however, we assume a multiple factor series to

30 THE AMERICAN NATURALIST [Vou. LIII

account for the facts, it is evident that, assuming all the
factors involved to have equal value, we must have an F,
type that is very nearly intermediate as regards its per-
centage condition. This we have not observed in any of
our series. The F, type, while being generally inter-
mediate, is not so as regards its hulled condition, for it
always contains fewer hulled kernels than hull-less.
Thus, so far as the percentage relation is concerned, we
do not have a strict intermediate. To be sure, there is a
reduction of the multiple-flowered spikelet and other
changes which cause the F, type to appear as an inter-
mediate.

With the usual multiple factor hypothesis assuming
ordinary segregation, there must be a larger number of
individuals ranging from 30 to 70 per cent. than we have
at the extremes. With series 379 and 202 we do not have
any indication of such a condition. On the other hand,
there is a slight suggestion that series 382 does tend
more nearly to a frequency distribution such as would
usually be expected with the ordinary multiple factor
hypothesis. When the third generation distribution of
series 202 is observed (Correlation Table V) it is ap-
parent that there is more of a tendency to pile up nearer
the lower values. When the size of the classes is doubled
a decided skew curve is obtained with the mode at class
0-9.9. As stated above, the seed sown to obtain the plants
used in this distribution was selected from plants of
high, low, or medium value, and this may influence to
some extent the type of distribution. Yet, when one ex-
amines the percentages of the plants used as parents, it
is apparent that they are fairly evenly distributed. If,
as suggested above, the nearly dominant primary factor
pair influences the hulled or hull-less condition and the
other factors influence the hulled condition of plants het-
erozygous for the primary factor, then we would expect
a piling up near the lower values.

With series 382 there is a tendency for both the third
and fourth generation percentage distributions to be

No. 624] INHERITANCE OF HULL-LESSNESS 31

grouped around the middle classes. This is especially
- true with regard to the fourth generation, especially
when the size of the classes is doubled.

The results of the different series are rather conflicting
and it does not seem possible at pregent to explain all of
them on a simple multiple factor hypothesis. It seems
quite possible to explain series 382 on this basis (except
the first generation) but the other types do not at present
seem capable of such an explanation.

The distribution in Table II, which is the third gen-
eration of a cross between Danish Island and Hull-less,
is skewed much the same as for the third generation of
series 202. No doubt for these series there is some dis-
turbing factor which causes such distributions and more
data will be needed before a suitable explanation can be
found to fit all of these cases. It may be that, since in
crosses between two hulled sorts we have found some
hull-less spikelets, we have combinations such that there
is a tendency to produce an excess of hull-less kernels.
This would influence the type of distribution consid-
erably.

At first one might assume that those individuals nearer
the lower part of the distribution were like the F, types,
however, from all the plants tested where the percentage
of hulled kernels has been low the frequency distribution
of the percentage of hulled kernels from the heterozygous
plants has been low in general and has not ranged from
very low to very high, as would be the case with seed from
F, plants. These facts would help support the statement
just made, which is to the effect that it is possible cer-
tain crosses tend to produce an excess of hull-less kernels.

SUMMARY

From the results presented it is evident that hull-less-
ness exhibits a simple Mendelian ratio of 1 hulled, 2 in-
termediate, 1 hull-less.

The intermediates show all gradations of hull-lessness
from those nearly hulled to those nearly hull-less.

-.

32 THE AMERICAN NATURALIST [Vou. LII

The percentage of hulled kernels on the heterozygous
plants seem, to indicate to some extent the percentage of -
hulled kernels on the heterozygous offspring.

No matter what percentage of hulled kernels is present
on the heterozygous individual, it tends in general to
produce a 1:2:1 ratio.

The hulled and hull-less kernels from intermediate
plants reproduce similar 1:2:1 ratios.

ENVIRONMENTAL REACTIONS OF
PHRYNOSOMA?

A. 0. WEESE

UNIVERSITY OF New MEXICO

I. INTRODUCTION

le General Distribution.—The horned lizards, more
familiarly known as the ‘‘horned toads,’’ of the south-
western portion of the United States and the northern
states of Mexico form a very distinct group of the family
Iguanide. Unlike most other comparatively large rep-
tilian genera, this particular genus (Phrynosoma) is
limited to a very special environment, and it is only in
a region of relative aridity that these animals find a
favorable habitat. Within the limits set by the above
condition the specific habitats of the various species and
varieties of the genus vary greatly, ranging all the way
from the extreme aridity and great heat of Death Valley
in southern California (Phrynosoma calidiarum Cope)
to the comparative moisture and cold of the northern
Rockies (Phrynosoma douglassti Bell and varieties).
The species especially discussed in this paper are all
found in the Southwest, under varying environmental
conditions.

Phrynosoma modestum, the specimens of which were
taken near Albuquerque, New Mexico, close to the lower
edge of the ‘‘mesa’’ or clinoplane region, at an altitude
of about 1,700 meters, is distributed throughout New
Mexico, and to a certain extent in the adjoining states,
wherever conditions are similar to those in the above
typical habitat. The rainfall here averages about 30 cm.
annually, while the yearly evaporation from a free water
surface is in the neighborhood of 200 em. The soil is

1Contribution from the Zoological Laboratory of the University of
Tlinois,

33

34 THE AMERICAN NATURALIST [Vou. LII

rather loose and friable, consisting principally of ‘‘Ti-
jeras fine sandy loam’’ and containing, near the surface,
a relatively large proportion of fine angular gravel and
wind-blown sand. The color is a yellowish or yellowish
brown. The vegetation is sparse, consisting of scattered
grasses, Chrysothamnus, Gutierrezia, Salsola, Yucca, ete.
This species is not found in the adjoining valley of the
Rio Grande, nor in the mountains (Sandias) which border
the ‘‘mesa’’ on the east (2,200 meters and above), where
moister conditions prevail. In the mountains the rain-
fall is probably twice as great, on the average, as on the
‘‘mesa,’’ although accurate data are not available, and the
evaporation is much less, due to the lower temperatures
which prevail. In the valley the water table is very near
the surface of the soil (actual soil surface or above to
5 meters below the surface). Standing water is not found
on the clinoplane except after very heavy rains, which
sometimes fail for months.

Phrynosoma douglassti ornatissimum, specimens of
which were obtained with the above, has a much less re-
stricted habitat, both locally and regionally. It is dis-
tributed over a great deal of the eastern slope of the
Rocky Mountains, even as far north as Canada, and,
locally, extends into both of the regions described above
as bordering on the clinoplane. It is, indeed, more abun-
dant in either of these than in the clinoplane region be-
tween, indicating that the determining factor in the dis-
tribution’ in this case is similar in the lower valley and
on the mountain side. As mentioned above, the aridity
of these two regions is much less than that of the clino-
plane. ‘The soil differences are also marked, in that the
moister soils are more dense and contain more humus,
derived from the more abundant vegetation. However,
the variation in both regions is very great, from heavy
clay to fine sand in the valley and from native rock to fine
sand in the mountain.

Phrynosoma cornutum does not occur in the same local
area as that occupied by the species previously mentioned,

No. 624] REACTIONS OF PHRYNOSOMA 35

although it also is of wide distribution. ‘This species is
found throughout Texas and eastern and southern New
Mexico, and has been reported from Nebraska, Arkansas,
ete. In general, it appears to inhabit regions in which
the mean summer temperature is slightly higher than that

Pies A aaa

me 4 er
`

f

Map showing the approximate geographical distribution of the species discussed
in this paper.

required by the other two species. The specimens here
considered were obtained at Alamogordo, in the Otero
Basin, New Mexico, where the mean temperature is higher
by about 5° C. than at Albuquerque.

2. General Habits.—The general habits of the three
species here considered are much the same, so no sepa-
rate description will be attempted. The following dis-
cussion will apply, perhaps, more accurately to Phryno-
soma modestum than to either of the other species, but
will, in general, be true of all. They are not, essentially,
heat-loving animals, although tolerant of desert condi-

36 THE AMERICAN NATURALIST [Von. LIH

tions. They are found more abundantly during the
earlier summer months, and during the autumnal rainy
season, when the aerial temperature does not exceed 32°
C. During these periods the animals move about actively
all day, spending the night in protected nooks under
vegetation, in the burrows of other animals, or buried
beneath the surface of the soil. As the daily maximum
temperature becomes greater they are to be found only
in the early morning and in the later afternoon when the
heat is less intense. During the heated part of the day
the lizard is at rest, almost if not quite buried under the
superficial layers of the soil. This position is reached in
a characteristic manner. "The snout is directed down-
ward and moved rapidly from side to side, the body ex-
tremely flattened, while the legs take part in a rapid hori-
zontally clawing movement. ‘The net result of this series
of movements is to cover the animal with the loose soil,
the depth varying according to the temperature, the char-
acter of the soil, and other external conditions, as well as
the individual. The same method of burrowing is em-
ployed in preparation for hibernation, when the animal
may bury itself under several inches of loose soil. In at-
tempting to escape from enemies, other lizards have been
observed to dig in a similar manner, and it is probable
that Phrynosoma also escapes in this way.

3. Food Relations.—The food consists of various in-
sects with which the animals come into contact, ants being
more readily eaten by the smaller individuals and beetles
(Eleodiini) forming a considerable portion of the diet
of the larger ones. No food is taken unless it is living or
at least moving. Sand grains set in motion by a heavy
wind or otherwise are often snapped up, and sand grains
are accordingly found in the feces.

4. Water Relations.—None of the species of Phryno-
soma have been observed by the writer to drink water,
and it is doubtful if water, independent of that contained
in the insect food, is ever ingested. Many individuals are
found in situations where there is never any standing

No. 624] REACTIONS OF PHRYNOSOMA 37

water except after the very infrequent heavy rains. Very
little water is excreted ordinarily, as when fed on ants,
beetles, etc., the feces are eliminated as a dry mass con-
taining practically no water, and the urine is composed of
an equally dry mass largely made up of crystals of uric
acid. When fed on a moist diet, such as grasshopper
nymphs from a moist habitat, the feces become softer and
are often accompanied by a considerable amount of muci-
laginous liquid. The urine, however, remains as usual.
The idea that the excretion of waste nitrogen as uric acid
is an adaptation on the part of the Reptilia for life in
arid regions is well borne out by the conditions in these
animals. Urinary analyses made by the writer in the
laboratory of physiological chemistry of the University
of Illinois give the following results (1917b):

onstituents Milligrams per gram
SEOUL iron (i A ccc eee s cee 260.0
SIND MORON oe ns tac ees uote 1.4
Pred neiroren OLOA Ga a. 0.0
EET T T: MEREEN TD i aces ese Oe <t 765.0
Uris Mole, Mitron sy ee es cw kc 255.0
Bara Silas ara ico ds EP eau ga mopman wate” 87.5
Phosphorus (as! PO), 205) 308 STO. Og OAS 3.5

It will be observed that uric acid accounts for prac-
tically all of the nitrogen contained in the urine and that
urea is entirely absent. In this respect the urine of the
horned lizard differs from that of the aquatic and semi-
aquatic reptiles, which contains a considerable amount of
urea, as does that of birds, another group in which the
uric acid content is high.

5. Reproduction.—It is in connection with Phrynosoma
cornutum that the long-disputed question as to the vivi-
parity or oviparity of the members of this genus may be
opened again. Cope (1898) states that Phrynosoma is
oviparous, which is denied by Ditmars (1908) and Watson
(1911), the latter of whom bases his statement on obser-
vations of P. douglassii. On July 5, 1917, some twenty
specimens of P. cornutum were received at the vivarium

38 THE AMERICAN NATURALIST [Vou. LIII

of the University of Illinois from Alamogordo, New
Mexico, and placed in a sand-bottomed wire screen cage.
On July 7, between 11 a.m. and 1 p.m., twenty-three eggs
were deposited in the sand on the bottom of the cage.
The eggs were about 1 cm. in length, ovoid in shape, and
covered with a grayish-white shell of leathery texture.
Some were opened and found to contain living embryos
of about 2 mm. length. Several times thereafter, during
a period of two weeks, eggs were found in the cage, always
lots of about twenty. The deposition of the eggs was
never observed. None of the eggs hatched, although liv-
ing embryos were found in eggs opened a week after
deposition. Such embryos were about 6 mm. in length.
P. douglassvi has not been observed to lay eggs, although
a few eggs of P. modestum were discovered in the cage
in which these animals were kept. These were found in
small numbers only and differed from those just de-
scribed in being light yellow in color and having no
leathery shell. They were probably abortive. As the
observations of Watson and Ditmars appear to be well
founded, it is possible that the genus is divided with re-
spect to the retention or deposition of the eggs, or that
in the same species different conditions may alter the
length of time the egg is retained in the maternal body, |
as is the case among the adders.

II. ENVIRONMENTAL FACTORS

As has been concluded (1917a), it is dangerous to as-
cribe to any one factor or group of factors the supreme
rôle in determining the seasonal or general distribution of
a species. These factors are certainly not the same for
all species even in the same environment, and before defi-
nite conclusions can be drawn a careful analysis of the
habitat must be made, and experimental data must be ob-
tained as to the reactions of the animals in gradients in-
volving the factors capable of variation. Unfortunately,
it is not possible or practicable to construct effective

No. 624] REACTIONS OF PHRYNOSOMA 39

gradients involving all environmental conditions, and in
such cases we must rely on careful observation and
analysis, Such a review as has just been given of the
habitat and habits of the horned lizards may indicate to
us the probable external conditions variations of which
are of importance in the daily and seasonal life of the
individual and of the species. The following are the most
apparent of such external conditions: -
1. Temperature. ;
(a) Air.
(b) Soil.
(c) Maxima and minima.
2. Water.
(a) Relative humidity and evaporating power of
air.
(b) Soil moisture.
(c) Food in relation to its water content.
3. Soil.
_ (a) Texture as influenced by
1. Composition.
2. Moisture content.
3. Vegetation.
- ¢b) Color.
4, Food.
(a) Character.
(b) Abundance or scarcity.
5. Light.
(a) Quality.
(b) Quantity.
(c) Rhythm.

In the natural habitat it is rare that one of the above
conditions varies without an accompanying variation in
one or more of the others; for example, a variation in
temperature of the air is accompanied by a variation in
the relative humidity and in the evaporating power of the
air, and may be followed by an alteration of soil tempera-
ture and soil moisture, as well as soil texture. Thus it
is difficult to consider these conditions separately.

40 THE AMERICAN NATURALIST [Vou. LIII

1. Temperature.—That temperature affects profoundly
the daily life of the animal and limits its activities is
shown by the relation of daily variation in temperature
to the change from diurnal to crepuscular habit and to the
burrowing activities initiated by high or low tempera-
tures. Minimum temperature is probably associated
most closely with the phenomena of hibernation. Ac-
cording to Bachmetjew (1901) the minimum winter tem-
perature which can be survived by hibernating insects
depends on the degree of elimination of water from the
tissues and the consequent lowering of the freezing point
of the body fluids. Tower (1917) states that in the case
of potato beetles those animals acclimated to desert condi-
tions (retention of water) are killed at higher tempera-
tures than those of a more humid climate. In the experi-
ments to be described gradients in air temperature and in
soil temperature (substratum temperature) were estab-
lished and the reactions of animals in such gradients were
recorded.

2. Water.—The water relation must always be impor-
tant in an animal adapted to arid conditions, even though
this relation may seem to be negative. As indicated by
the examination of excreta and observation of the water
relations of Phrynosoma it would appear that the ab-
sence of water as such would not have a limiting effect
on the distribution of the animals. It is probably neces-
sary, however, that a certain minimum amount of water
be supplied in the food, and that the evaporating power
of the air must not exceed a certain maximum for any
great length of time. It is to be doubted that any verte-
- brate may subsist indefinitely without some small water
supply in addition to metabolic water. As shown in
previous experiments (1917a), the reaction of Phryno-
soma in a gradient of the evaporating power of air is not
definite unless the gradient be very steep. Daily varia-
tion in the normal habitat is very large.

3. Soil.—The apparent importance of the burrowing
reaction in the life history of the members of this genus

No. 624] REACTIONS OF PHRYNOSOMA 41

points to a corresponding importance of the texture of
the soil. Evidently this must be such as to render the
success of the burrowing reaction comparatively easy, a
condition which is met only in soils of a low moisture
content, and little humus, containing a considerable
amount of loosely aggregated particles of sand or fine
gravel. In a heavy clay or loam it would be impossible
for the animal to burrow deep enough to get below the
zone of killing temperatures during hibernation. This
would also be impossible in a compact sod. Unfortu-
nately, the problem of the soil relation involves an ex-
tensive seasonal study which, so far, it has been impos-
sible to carry out.

While the color and markings of the animals vary with
the individual and the species, and the color of the indi-
vidual changes from time to time, it may be said in gen-
eral that the color of the horned lizard is very similar to
that of the soil of its normal habitat. Experiments of the
author and others have shown that high temperature,
darkness or high evaporating power of the air causes a
centripetal movement of the melanophoric pigment, while
the opposite conditions cause a darkening. Thus, in gen-
eral, individuals observed after a rain are darker in color
than at other times. The soil is also darker when wet,
which might lead the observer to suppose that the change
had taken place as a direct adjustment to the color change
of the substratum, while the actual cause is the change in
the evaporating power of the air. Within the limits of
the conditions of the habitat, variations in the evapora-
ting power of the air are the most potent factors in the
production of color changes. No direct connection be-
tween the color of the animal and that of the substratum
has been verified experimentally by the author. Redfield
(1917), in a recently published paper on the color changes
in Phrynosoma cornutum, has stated that there is a direct
approximation of the color of the animal to that of the
substratum, and that the light rays reaching the retina
form the stimulus for such changes. The mechanism for

42 THE AMERICAN NATURALIST (Vou. LIII

the approximation of the color of the animal to that of the
substratum is, according to Redfield, subordinate to the
daily rhythm of color change occasioned by changes in
light and temperature, and to changes brought about by
the emotional condition of the animal.

4, Food.—An adequate study of this factor would re-
quire much more extended observation than has been pos-
sible. Some suggestions as to the character of food re-
quired have been made above.

5. Light.—An estimation of the effect of light of vary-
ing intensity and quality in the natural habitat would be
very difficult, but it is probable that the relations of light
in the life of such animals have been greatly underesti-
mated. Experiments with a gradient of the color of light
are included here.

Ill. EXPERIMENTAL RESULTS

1. Air Temperature Gradients.—Two series of experi-
ments were performed in which air varying in tempera-
ture was passed across the experimental cage previously
(1917a) described. In the first series the air passing
across one third of the cage was heated to a temperature
of about 38° by being passed through coils immersed in
hot water, that passing across the next third was heated
to about 33°, while the remaining third was supplied with
air at about 29°. The air was unmodified except as to
temperature and the rate of flow was the same in each
case. Typical results of this series (Phrynosoma mo-
destum only) are shown statistically in Table I.

In the second series the air for the hottest third was
heated to a temperature in the neighborhood of 50°,
which is about the maximum soil surface temperature on
unprotected sand exposed to the direct rays of the sun.
This temperature was obtained by passing the air through
heated iron pipes. A medium temperature was obtained
by passing the air through coils immersed in hot water,
as above, while the lowest temperature was that of the

No. 624] REACTIONS OF PHRYNOSOMA 43

TABLE I
EXPERIMENT 34, SHOWING THE REACTIONS OF sbi modestum IN
AIR TEMPERATURE GRADI
Ten animals were placed in the cage, and observations of their position
taken at one-minute intervals. The temperatures taken at intervals along
the cage are indicated at the heads of the respective columns.

Temperatures

Minutes Experiment 344 Experiment 34b

Ny
©
°
wo
oo
°
oe
@
o
o
@
o
kd
w
°
nN
©
°

OCONoorwde

a
pe ee ee eet et DDD NNW NNN NRWWWWWWW PP Rp
PREP ROW WWW RPE PR ROR RP PP RDN NNW
OR EE ere aa
OS oa o oa o w oe O O Oe A PR WWW Ph PP Pe Be PP RB Co OO 00
AAAMWAOMIEE EPONA aa ;a kr OS oe
Te es a rs hm es nb te be

unmodified air, about 30°. These temperatures varied
somewhat in the various experiments, as shown by the
records, but were fairly constant throughout a single
experimental period.

The records of Experiments 34a and 34b show, for
Phrynosoma modestum, that the optimum air tempera-
ture is in the neighborhood of 35° or 36°. The graphic
records of Experiments 86 and 88 (P1. I) show similar
results. It will be noticed in the record of the former

P. cornutum P. cornutum, P. douglasii P. douglasii P. modestum P. modestum
substratum substratum substratum substratum substratum substratum
temperature temperature temperature | temperature temperature temperature,
ni
51° 50° 47-3954 430'43° 43° 360 28° |e lee [zr | 48° 48°46°40°27° 37° 37° 35° 27° 19° 47° 46° 45° 39° 31:
= = pe) ar

PLATE I. Illustrating the reactions of Phrynosoma in gradients of air
temperature and substratum temperature. Experiments No. 45, 66, 101, 26, 12,
28, 86, 88, 82, 85, 78, 81.

In the chart. each section between the numbered scales represents the
record of a twenty-minute experiment, the distance between the scales s repre-

ge, an

found in the graphic records o rtain Aadamal pei that the animal
pie neh to burrow at the Haah ‘tals
Con , 4. e., experiments in which all portions of the cage were at the
same base ina oe were carried out in all cases, but the regular curves obtained
have been omitted nip save spac
E r the fei two minutes the animal was comparatively
quiet, and after thet close of this period moved toward the hot end of the cage,
o return immediately, and then attempt o burrow
minute the animal again moved toward the higher temperature and again bur-
rowed. Thereafter the movements ie of greater amplitude but less frequent,
until the erp finally came to rest near the cooler end of the cage, where it
remained until the end of the priate period.

7 81
odestum P. modestum P. "o P. douglas P. comutum P. cornutum
A < 3

siele, _lolebe| i i et aee iTe ziele]
P Piet i: 2 gate mee
= 2 ‘ Ss - LO-
: l 2 : s ; S E

periment 66 shows a record of almost the same character, except that all
movements were of lesser amplitu
Experiment 101,—Phrynosoma dougtassi here remained for the greater part
of the time at a temperature sa bout 88°, making infrequent excursions into
the region of lower te emperatu

“periment 26.—Same siah above. The temperatures here were
higher, and the animal avoided igh pana temperatures.
eriments 12 and 28 show results similar to the two just preceding, in one

æpe
the a miresei of low temperatures and in the other the avoidance of higher
Pe
t 86.—This is the record of the movements of a very sluggish
aaa "whieh gh tet twice at @ temperature a little above that chosen
by other members of t ae nee
Experim wed alternate periods of rest at an opti-
mum temperature ‘aad Fade ‘involving incursions into both temperature ex-
tremes represented in the a radie
Experiments 82 and 85 are paN examples of the type a Delmas there is
great activity, but very short stays in the unfavorable tempera
@wperiment 78,—This record shows avoidance and turn wis et from the
higher temperatures. The ithir the animal penetrated into the high tem-
. Þerature area before turning, the longer was the subsequent inactive period in
e.

ent 81,—This animal was very sensitive to the sya temperatures
and never reached the hot on although very active at tim

46 THE AMERICAN NATURALIST [Von. LIII

that the animal burrowed, first at a temperature of about
38° and later at a slightly lower temperature (indicated
by the circles in the first and sixth minutes of the record).
This burrowing reaction was found to take place very
often, throughout the whole series, usually at the upper
limit of the optimum temperature range. This agrees
with the phenomena observed in the field, of burrowing
as the air temperature rises in the course of the day.

Phrynosoma douglassii, as shown in the graphic records
of Experiments 82 and 85 (P1. I), seems to choose a some-
what lower temperature, between 30° and 35°, although
there is a considerable amount of individual variation.

Phrynosoma cornutum, the behavior of which in the
gradient is illustrated by the records of Experiments 78
and 81 (P1. I), appears to show a preference for a tem-
perature slightly higher than that shown by the other
species.

2. Substratum Temperature Gradients.—For the pur-
pose of establishing this gradient the cage was placed in
a water bath so arranged that hot water flowed into the
latter at one end and cold water at the other, the water
being directed backward and forward beneath the cage,
and running out near the center, in such a manner as to
produce a gradient in the temperature of the cage bottom.
The temperature of the substratum was taken at intervals
along the edge of the cage by thermometers whose bulbs
were just covered by the sand in the bottom.

The statistical records of Experiments 43, 126 and 127
(Table II) show an optimum substratum temperature for
Phrynosoma modestum of about 40°, or about 5° higher
than the optimum air temperature for the same species.
In this species the response to changes of temperature of
the substratum is very definite, and by varying the tem-
peratures of the gradient, the animals can be driven re-
peatedly from one end of the cage to the other as the
temperature is raised or lowered. The lizards often bur-
rowed at or near the upper limit of the optimum tempera- .
ture, and, less often, at the temperatures below the

No. 624] REACTIONS OF PHRYNOSOMA 47

optimum. The graphic records of Experiments 10 and
28 (Pl. I) show similar results.

TABLE II
EXPERIMENTS 43 AND 127. SHOWING THE REACTIONS OF Phrynosoma
modestum IN A GRADIENT OF THE TEMPERATURE OF THE TUM
The method of recording is the same as that employed in Table I.

Temperatures :
Minutes Experiment 43 Experiment 127
45° 41° 36° 25° 40° 52°
4 2 4 2 2 6 2
2 2 4 2 1r 8 0
3 2 4 2 ta 8 0
a 2 5 1 1 8 1
5 2 5 ł 1 8 1
6 2 4 2 d 8 1
7 2 5 1 1 8 1
8 3 4 1 1 9 0
9 1 6 1 1 9 0
10 1 5 2 1 9 0
42 1 5 2 1 1*8 0
12 2 5 1 1 18 0
13 2 5 1
14 1 6 1
15 1 5 2
16 2 5 2
17 0 6 2
18 0 6 2
19 0 6 2
20 0 vi 1

* The individuals indicated by the italic numerals burrowed in the space
indicated.

The individuals of Phrynosoma douglassii gave prac-
tically the same figures for the optimum substratum
temperature. The graphic records of Experiments 101
and 26 indicate the behavior of this animal in the gradient.
Statistical records of the behavior of Phrynosoma doug-
lassii and Phrynosoma cornutum in this gradient were
not made, because of the size of the animals, which pre-
vented the introduction of any number into the cage at
the same time.

Phrynosoma cornutum, as silastrated by the records
of Experiments 45 and 66 (Pl. I), chose a higher sub-
stratum temperature than either of the other species,
averaging nearly five degrees above that shown by the
other curves.

48 THE AMERICAN NATURALIST [Vov. LIII

In summing up the results of the air temperature and
substratum temperature experiments (over one hundred)
in relation to those of the evaporation gradient previously
reported (1917a), it is found that the animals choose con-
ditions which are very near the normal conditions in the
usual habitat at the time of the greatest activity. These
conditions represent the optimum for the animals. For
example, as reported in a previous paper, the evapora-
tion optimum for Phrynosoma modestum appears to be
near 3 c.c. per hour, as measured by the standard atmom-
eter, which is very near the average outdoor evaporation
as observed in the natural habitat of the animal at the
season and at the time of day when the animal is most
active. If the temperature under such conditions be ob-
served, it will be found that the average atmospheric
temperature, 1 cm. from the surface of the soil, in the
sun, is in the neighborhood of 35°, and that of the surface
layer of the soil about 40°. These temperatures vary
greatly, of course, with other features of the weather,
such as air movements, clouds, ete., but the above figures
represent a normal condition. Of the variables men-
tioned here, substratum temperature has much the
greatest effect on the behavior of the animal.

3. Moisture of Substratum Gradient.— Although it was
impossible to establish and observe an effective gradient
in general soil conditions, several experiments were per-
formed on the direct effect of a soil moisture gradient.
The gradient in water content of the substratum was ob-
tained by placing a layer of torpedo sand saturated with
water on the bottom of one third of the cage, a mixture
of saturated sand and dry sand in the adjoining third,
and dry sand in the remainder of the cage. In none of
the species observed was any marked preference for any
portion of the cage exhibited. Soil moisture, as such,
does not seem to affect the movements of the animals,
although, in the natural habitat, the high evaporating
power of the air produces a considerable degree of tem-
perature difference between dry soil and wet soil by the

No. 624] REACTIONS OF PHRYNOSOMA 49

vaporization of the water from the latter. This differ-
ence was not reproduced under experimental conditions.
Typical results of this series of experiments are shown in
the graphs of Experiments 91 and 97 (PI. II).

2

112 116 118 91 97

P. modestum P. comtan P. mniam P. modestum P. g
lrioly|claly| JR|olyicigly] lrloly] |ely|_ |pimiwl_ |p |miw! |

T

I. Illustrating the reactions of Phrynosoma in gradients of wave-

Tiia FRN of light and moisture of substratum. Experiments 112, 116,
118, 91 and 97.

Experiment 112,—In this and the two experiments following, a letters R,

and V above the graphs represent the color of the light screen over

yellow
Eeperiment 116—This animal avoided the violet and even the blue very
markedly Pg entered the red only twice. The optimum seems to be in the
yellow an oe een

Evperiment phe idance of both the violet and the red is nary i
of this curve. This animal, however, did not avoid the blue and the orange, in
which it et a considerable amount of time.

Exp ents 91 and 97,—Here the letters D, M, and W refer to dry, medium
and wet tine of the cage. The graphs show no preference for either on the
part of the animal.

50 THE AMERICAN NATURALIST [Vou. LIII

4. Gradient in the Color of Light (Wave-Length).—
Although it would be difficult to estimate the effect of the
various light components in the natural habitat, a series
of light experiments has been included. For use as a
color gradient the cage used in the other experiments was
covered with an accessory lid composed of a series of six
equal strips of gelatine ray filter in the principal colors
(violet, blue, green, yellow, orange, red). Three forty-
watt electric lamps were placed above the cage within the
observation hood, so that the light was approximately
equally distributed throughout the cage, each sixth being
illuminated principally by rays of a narrow range of
wave-length.

Experiment 112 illustrates the movements of Phryno-
soma modestum in such a gradient. The longest rays
were avoided, as well as the shortest, although the animal
remained for greater lengths of time in the red section
than in the violet. The optimum seems to lie in the green
and the yellow.

Phrynosoma cornutum (Experiments 116 and 118, PI.
II) avoided both red and violet, with an optimum near
the middle of the spectrum. Phrynosoma douglassii did
not respond regularly and seemed little affected.

The color reactions are probably not as significant as
those involving some of the other factors here considered.
Direct sunlight in the arid regions contains a rather
larger amount of the light of the shorter wave-lengths
than elsewhere, and it is possible that the avoidance of
violet light as shown in these experiments is of signifi-
cance in explaining the avoidance of sunlight under cer-
tain conditions, but it is more probable that temperature
is the dominating factor in this reaction.

IV. Summary AND CONCLUSIONS
1. Of the temperature conditions capable of being
tested in the gradient, the temperature of the substratum
calls forth the most definite response. In addition to the

No. 624] REACTIONS OF PHRYNOSOMA 51

indication of an optimum by the movements of the animal,
definite motor responses of a specialized character (bur-
rowing) are made to certain temperature conditions just
above or just below this optimum. The temperature of
the air calls forth similar reactions but not as readily or
as definitely as that of the soil, the reaction to the former
being overshadowed by the response to the latter when a
difference exists. The temperature of the substratum is
evidently of very great importance in the daily move-
ments of the horned lizards, and probably plays an im-
portant rôle in the control of distribution. The tempera-
ture of the soil is probably also of great importance in
connection with the deposition and hatching of the eggs
in those species which are oviparous. The differences
between the optimum temperatures of the various species
considered are in the direction and of the magnitude of
the temperature differences normal to their respective
habitats. While the limits of temperature variation
favorable for the completion of the life cycle of the animal
could not be subjected to experiment of the type here
used, it is evident that at least the minimum is of great
importance in connection with the phenomena of hiber-
nation, and the maximum is probably of similar impor-
tance in relation to the estivation which takes place more
or less regularly. ;

2. In the gradient of the evaporating power of air
definite responses were obtained only in the case of one
species (Phrynosoma modestum), and here only when the
gradient was steep. The daily and seasonal variation
in this factor is very large in the natural habitat. The
reactions of the animals to temperature changes act in
such a way as to prevent the exposure of the organism to
excessive desiccation. The effect of soil moisture is felt
indirectly, through the alteration of the temperature and
the texture of the soil, the latter of which is important
in relation to the burrowing habit. It is probable that
there is a certain minimum water content of food, below
which the animal can not survive. This must be very

pro- THE AMERICAN NATURALIST [Vor. LII

low, however, considering the character of the normal
food. The excretion of water is reduced to a minimum
by the character of the nitrogenous excreta, which are
almost exclusively in the form of insoluble uric acid.

3. An important factor in the distribution of these
animals is the texture of the soil, which must be suitable
for burrowing, as this is the reaction of the animal to
unfavorable conditions generally, and specifically to tem-
peratures inducing hibernation and estivation. The soil
texture is affected adversely by increases in moisture con-
tent, and by increases in the amount of vegetation present.
The color of the soil is probably important from the stand-
point of invisibility and it is probable that there is some
degree of approximation of the color of the animal to that
of the substratum. It is difficult to see how this fact
could be of much use to the animals, especially in the case
of such profusely armored species as Phrynosoma
cornutum,

4. The rôle of light in the daily and seasonal life of the
horned lizards has not been shown, although they are posi-
tively phototactic and avoid extremes in a color gradient.
The optimum in this gradient lies in the green and in the
yellow. This may be correlated with the predominant
colors of soil and vegetation in the natural habitat.

V. ACKNOWLEDGMENTS AND BIBLIOGRAPHY

It is with great pleasure that I express my gratitude to
Professor Victor E. Shelford, of the University of Tli-
nois, for his constant inspiration and ready assistance
rendered in the performance of this work.

LITERATURE CITED

Babcock, S. M.

1912. Metabolic Water: Its Production and Rôle in Vital Phenomena.

Wisconsin Research Bulletin, 22: 87-181.

Bachmetjew, P.

1901. Experimentelle entomologische Studien. Leipzig.
Bailey, Vernon.

1905. Biological Survey of Texas. No. Am. Fauna, 25.
Cope, E. D.

No. 624] REACTIONS OF PHRYNOSOMA 53

1898, The Crocodilians, Lizards and Snakes of North America. Re-
port U. S. N. M., 1898: 153-1270.
Dice, L. R.
1916. pnn of the Land Vertebrates of Southeastern Washing-
Uni. Cal. Publ. Zool., 16: 293-348.
Ditmars, R. Pies
1908. The Reptile Book. New York.
Girard, Charles A.
1853. A Monographie Essay on the Genus Phrynosoma. Stansbury’s
Exped. Gt. Salt Lake, 354-365.
Grinnell, Jos.
1917, Field Tests of arson Concerning Distributional Control.
Am. Nar., 51: 1 5-128.
Herrick, C. L., Terry, Fötin, Ki Herrick, H. N.
1899. Notes a Collection Ne; tives from New Mexico. Bull.
Sci. oe Dennison Uni., 11: 117-148.
Nelson, J. W., Holmes, L. C., and Eckmann, E, C.
1914. Soil Survey of the Middle Rio Grande Valley Area, New Mexico.
S. D. A., Advance Sheets, Field Operations, Bureau of
Soils, 1912.
Parker, G. H.
1907. ee of Light and Heat on Melanophores. Jour. Ex, Zool.,
: 401-439,
Redfield, ree C.
1917a. The Pae of the Melanophores of the Horned Toad.
Proc. Nat. Acad. Sci., 3: 3.

1917b. The bose of the Melanophore Reactions of the Horned
Ibid.

Toad
Pritchett, Annie E.
Some Experiments in Feeding Lizards with Protectively Col-
ored Insects. Biol. Bull., 5: 271-287.
Ruth, E. S., and Gibson, R. B.
1917. Disappearance of the Pigment in the Melanophores of Philip-
pine House Lizards. Phil. Jour. Sci., Sec. B., 12: 181-190.
Shelford, V. E.
1911. Physiological Animal Geography. Jour. Morph., 22: 552—618.
1912, ak ogiċal Succession. V. ae ts of Physiological Classifica-
on. Biol. Bull. 1-370.
1913a. Aina PRAA in Temperate America. Geog. Soc. Chi.,

1913b. The Reactions of Certain Animals to Gradients of Evaporat-
Power of Air. Biol. Bull., 25: 79-120.
1915. Principles and Problems of Ecology as Illustrated by Animals.
Jour, Ecol., 3: 1-23,

Tower, W. L.
1917. Inheritable Modification of the Water Relation in the Hiberna-
tion of Leptinotarsa decemlineata, Biol. Bull, 33: 229-257.
Visher, S. S.
1916. The Biogeography of the Northern Great Plains. Geog. Rev.,
1: 89-11

54

Watson, J. R.
1911. A Contribution to the Study of the Ecological Distribution of

1912.

THE AMERICAN NATURALIST [Vou. LIII

the Animal Life of North Central New Mexico with Especial
Attention to the Insects. Rep. Nat. Res. Survey, N. M., 1

57-117.
Plant Geography of North Central New Mexico. Bot. Gaz.,
54: 194-217.

Weese, A. O.
1917a. An Experimental ee of the Reactions of the Horned Lizard,

Weinzirl, Jo
0

Phrynosoma modes Gir., a Reptile of the Semi-desert.
Biol. Bull., 32: AN (Bibliography).
an Urine of the Horned Lizard. Science, N. S., 46: 517-518.

Driana from Water Surface at Albuquerque, New Mexico.
ull. Hadley Climatol. Lab., 3: 10: 1-14.

MIGRATION AS A FACTOR IN EVOLUTION: ITS
ECOLOGICAL DYNAMICS, II.

CHARLES C. ADAMS

PROFESSOR OF Forest Zootogey, THe New York State COLLEGE OF
Forestry AT SYRACUSE UNIVERSITY

Ill. Tue MIGRATIONAL Factors IN EVOLUTION

1. Introduction

From the preceding discussion of the principles of ani-
mal activity which underlie their behavior, attention is
now directed in greater detail to suggestions for their
application to migration. Emphasis is placed upon those
relations which show the main causes of stress, the cycles
of circulation caused by diversity, and the interaction,
equilibrium, and adjustment operating between the vari-
ous systems. Ihave not attempted to go into detail on the
quantitative relations, although there is much physical
and some ecological data, already organized, which illus-
trate the method of application. There is, however, but
little quantitative distributional data which are at pres-
ent available. The elaboration of this phase is urgently
needed. Limiting factors retard and prevent the migra-
tion and diffusion of animals; these are the ‘‘barriers’’
so frequently mentioned by students of geographical dis-
tribution. As previously mentioned, two major systems
or agencies are involved in this process, the animal and
the environment. The locomotor activity of the animal
is a phase of- its general responses. The migrations of
most animals are therefore not different, in any impor-
tant features, from the ordinary daily life of the animals;
that is, migrations are incidental and included within the
ordinary responses. Anticipating somewhat, and speak-
ing broadly, if animal responses are of evolutionary value

55

56) ` THE AMERICAN NATURALIST (Vou. LII

so must be the migratory ones. In detail there are in-
numerable animal peculiarities which influence migra-
tion, depending on the stage of development of the ani-
mal, its physiological and ecological conditions and char-
acteristics, and the status of its environment. The large
number of factors involved in this is no doubt an impor-
tant conservative influence and checks the speed of in-
teraction.

The word migration is used in several senses, so that
for our purpose it is now necessary to define this more
definitely. By migration is here meant the movement of
animals from one place to another, and this includes, not
only the causes and conditions of their migration, but
their methods as well and the immediate result upon the
animal. If all migrating animals perished at the end of
their journey the study of its influence would be rela-
tively simple.

In deference to those who are mainly hiat i in the
animal and who are less interested in the environment it
has been customary in many zoological writings to dis-
cuss animals first, and their environment later. But as
zoology progresses and as explanations are resolved
more and more into the sciences upon which it rests,
greater and greater prominence is given to the physical
causes and conditions of the environment. ,Viewed
broadly, zoology should be made to fit into the general
world system in such a manner as will best aid in under-
standing it, irrespective of our traditional habits of mind.
For this reason this phase of the discussion will begin
with the environment, as a factor in passive migration or
transportation, and will lead up to the animal as a factor
in its own migrations.

In the orientation of the major features of the world
Powell recognized: the atmosphere, the hydrosphere, the
lithosphere and the biosphere. These self-explanatory,
relatively homogeneous, physically and mechanically dis-
tinct, and interacting systems, furnish the medium in
which animals live and perform their migrations. We

No. 624] MIGRATION A FACTOR IN EVOLUTION 57

may look upon these three physical systems as a result
of existing temperature conditions of the earth. Increase
the temperature to a certain amount and the atmosphere
would be rarefied, the hydrosphere would disappear into
the atmosphere as water vapor, and the solid earth would
become fluid. Or, reverse the process, beginning with its
present state, and should the earth cool progressively,
the hydrosphere would become frozen to the solid phase
and atmosphere would likewise be transformed to
the solid state, and all these systems would become one.
The present resolved and differentiated state is thus de-
pendent on the present temperature conditions. The re-
lation of equilibria between these three systems is one of
the major problems for the application of the phase rule,
and their methods of interaction is an unlimited field for
the application of Bancroft’s law, and both of these are
of the greatest importance to all concerned with the gross
environments of organisms.

In view of the dominating influence of temperature, we
must not overlook the fact that temperature is only one
of the essential conditions of life. It is important to ob-
serve that the present stratum of the earth’s surface
where organisms live is a remarkably narrow one, and
only moderate departure above or below the condition in
this stratum at once becomes limiting factors to organic
activity. Chamberlin (’06, pp. 1-2) states this impres-
sively as follows:

The narrowness of the range to which temperatures must be confined
to permit progressive organic and intellectual evolution takes on its true
meaning only when we recall that the natural temperature range on the
earth’s surface is sixteen times as great as this, while that affecting
the solar family is at least sixty times as great. For a hundred million
years, more or less, this narrow range of temperature has been main-
tained quite without break of continuity, unless geologists and biologists
are altogether in error in their inductions.

The maintenance of such a dynamic system of equilib-
rium of the environment and of the organism, and the in-
ertia of their systems—the tendency to continue or per-

58 THE AMERICAN NATURALIST [Von. LIII

sist in a given state—may well cause wonder and stimu-
late thought.

In the following analysis of the larger units which in-
fluence migration, those agencies will be used which serve
as the basis for the smaller systems of action, and some
of their main cycles of activity and methods of interac-
tion will be indicated briefly.

2. Atmospheric Agencies in Transportation

The instability of the air, its numerous cycles of activ-
ity or circulation, hourly, daily, seasonally, annually, and
those of longer duration, furnish an agency which has
transported animals from one locality to another for
ages. Gentle breezes carry small animals, while violent
tornadoes carry larger ones. Small eggs, desiccated
rotifers, entomostraca, and other small aquatic animals,
have been transported long distances by the wind, and
have thus found many favorable habitats, otherwise not
available to them. The wind, reinforced by streams,
even temporary ones, has transported animals long dis-
tances, as have the waves of the sea and inland waters.
The winds, supplementing the flight of animals, have car-
ried them thousands of miles beyond their normal range,
as in the case of birds and insects. A vast literature has
grown up recording the details of these findings, and yet
about the only evolutionary conclusion which can be
safely drawn from the multitude of facts is that by these
processes animals have tried out and acclimated them-
selves to a vast number of isolated habitats which have
tended to give them a varied and widespread range, and
to that degree it has aided in their perpetuation.

The most definite evidence of atmospheric influence in
evolution is perhaps the direct influence of climate and
of climatic changes. Fortunately, for our present pur-
pose, and mainly through the researches of Chamberlin
(1897-1901) and Huntington (’15, pp. vi-vii), secular
climatic cycles have been investigated. Chamberlin has
related these intimately to the changes in the hydrosphere

No. 624] MIGRATION A FACTOR IN EVOLUTION 59

and lithosphere, and he has indicated their modes of in-
teraction in a strictly dynamic manner. He shows that
during a period of land elevation and mountain forma-
tion, with cold, dry, diverse climatic differences and zonal
arrangements on land and with a deepening of the sea,
these conditions tend to change progressively toward a
moist, warm, uniform and tropical condition, which is
related to the land equilibrium developed during base-
level on land, and a marine condition of extensive shallow
seas. The process of adjustment to these strains beauti-
fully illustrates Bancroft’s law. It is not an accident
that mountains are centers of origin and dispersal of ani-
mals, nor are they solely refuges where endemic forms
escape the competition of the lowlands. Mountain re-
gions in their elevation subject whole populations to
severe climatic and other stresses of many kinds, depend-
ing on the physical and vegetational diversity of the
region, and doubtless thus many animals become extinct,
while others as individuals or as a race become acclimated
to the new and changed conditions and thus survive.

It seems strange that, although dynamic principles are
shown in almost ideal form in the unstable air, yet, as a
whole, this phase of science seems to be somewhat back-
ward in the formulation of the ideas of processes, so that
their greater successful application is seen in geology. It
appears that the reciprocal dynamice relations which exist
between barometric lows and highs (both temporary and
permanent) makes them dynamic centers of action (Fas-
sig, 99) in a cycle of circulation and adjustment to stress.
This idea is one which may profitably be extended to the
interpretation of successive phases in the establishment
of climatic dominance. The change from the Ice Age to
that of the present, and the accompanying change of
storm tracks (Adams, ’09, pp. 45-46) are comparable to
the seasonal change from winter to summer dominance,
while passing through the transitional March weather
stage. Furthermore, the summer and winter dominance
are dynamic equilibria established by a balancing among

60 THE AMERICAN NATURALIST (Vou. LIIT

the various highs and lows (Adams, ’15a, pp. 69-71).
These transitional periods illustrate Bancroft’s law in
the process of establishing new relative equilibria. The
dynamic centers are to be looked upon as concentrating,
transforming and radiating centers, whose recognition
. and cycles of activity are an essential part in the applica-
tion of Bancroft’s law to the development of atmospheric
equilibria.

3. Hydrospheric Agencies in Transportation

The waters of the earth are more dense than the air
but are yet quite mobile, and undergo relatively rapid
cycles of circulation, both in the sea and in inland waters.

(a) Marine.—The great currents of the sea, the tides,
and the wind-formed waves, are very active agents in the
transportation of animals. Not only are marine animals
extensively transported, but also, in the long run, large
numbers of land kinds, as the animals on oceanic islands
testify, as shown by Wallace. And just as the processes
of erosion operate upon land, and tend to reduce such
areas to sea level, so the sea itself possesses its own
cycles of transformation of its bottom and shores, tend-
ing to flatten them out to the equilibrium of the deep sea
floor, transporting materials and redistributing them in
response to its stresses, eroding here, depositing there,
and always making changes in the conditions which not
only transport animals, but as well, by the migration of
the physical conditions, lead animals from one locality
to another. Land animals are largely influenced by the
surface conditions, while the marine ones are largely by
sub-surface conditions.

(b) Inland Water Bodies.—Bodies of inland standing
water, in their broad features, are smaller editions of the
seas, as far as their waves, circulation and transport-
ing powers are concerned. Their chemical character,
whether fresh or saline, has more influence upon animals
than the mechanics or their methods of circulation. The
most marked influence of the inland waters is their rela-

No. 624] MIGRATION A FACTOR IN EVOLUTION 61

tively small area, isolation, even though they may over-
flow into streams. Inland waters are in general rela-
tively ephemeral in character compared with marine
waters, because with progressive erosion of the land they
tend to become extinct through deposition and ultimate
drainage.

(c) Running Waters.—The transporting power of
running water is easily evident. The constant direction
of flow, its duration (as some drainage lines are of ex-
treme antiquity), and repeated transportation, have sub-
jected animals again and again to new conditions, and
carried them to new localities. Streams transport both
land and aquatic animals and by their persistence, activ-
ity, and the thoroughgoing fashion in which they work over
the land surface, are one of the most powerful agencies
of transportation. Streams undergo changes depending
on the dynamic status of the stream. The greater stress
to which the stream is subjected by uplift, the greater its
velocity and its relative transporting power, and the
nearer it erodes to base-level, the less current and relative
transporting power it possesses. Most animals counter-
act the transporting power of the stream by definite re-
sponses to the current, and thus maintain their position
and are not carried away.

4, Lithospheric Agencies in Transportation

The lithosphere includes the solid earth, which to the
ordinary mind is the ideal of stability. The transport-
ing power of the solid is, however, usually at a very slow
rate, but this is not always the case, because of the sud-
denness of fracture. The solid ice of the glacier moves
slowly and yet travels long distances, but usually does
not transport an abundance of animals. Avalanches
move with greater speed, but they operate in rather
limited areas. Landslides transport, slowly or rapidly,
large masses of land containing animals. All of these
processes are dominated by gravity, and tend to trans-
port animals from a higher to a lower altitude. Perhaps

62 THE AMERICAN NATURALIST [Vou. LIII

the most powerful motion of the solid earth is seen in the
crustal movements, associated with the cycle of isostacy,
which elevates and depresses the surface of the land in
relation to sea level. In this is seen an essential condi-
tion which has made all land life possible, because with-
out such movements all the land would have been washed
into the larger dominating sea. The great land eleva-
tions, such as those which produce plateaus and moun-
tains, have transported whole faunas, covering thou-
sands of square miles, upward, and have subjected them
to great stress, through long periods of time. Such ele-
vations as arise in a region unfavorable to animals, may
improve them, as in the case of high mountains, rising on
a dry desert, but often such elevations, which are de-
partures from the favorable thin surface stratum, are
in the direction of unfavorable conditions and of limiting
factors. Broadly speaking, depressions below sea level
are similarly limiting to marine organisms, and these
have operated on a magnificent scale. The mountain
tops, like the deeps of the sea, are relatively animal
deserts, both are extreme departures from the conditions
which are normal to most animals.

The most rapid physical agency in the transportation
of animals on land is the influence of running water and
that of the wind. These forces operate in short cycles
and intensively, in contrast with the movements of the
solid earth.

Voleanic activity has probably been only a minor factor
in the transportation of animals, although in a secondary
way, in conjunction with other agencies, as currents of
water, porous materials buoyed by air, may act as a raft
in their transportation. But indirectly by building moun-
tains, islands, ete., it has had an influence similar to that
of the crustal movements of the earth in forming new
habitats, and has thus had a powerful effect.

No. 624] MIGRATION A FACTOR IN EVOLUTION 63

5. Biospheric Agencies in Transportation and
Migration

(a) Plants.—The relatively sedentary and rooted con-
dition of plants caused Cope to aptly call them ‘‘earth
parasites.’’ With this stable habit and the unstable en-
vironment, rooted plants have been forced to develop a
line of fracture, as it were, between themselves and the
environment, which has permitted them, by their excep-
tional powers of dispersal, to spread rather freely at
some stage, and to thus scatter over much of the avail-
able surface of the earth. As far as the actual move-
ments of plants are concerned, unaided by winds, wat-
ers, and animals, but solely by growth and similar move-
ments, plants have probably had but relatively small
influence upon animal transportation, although second-
arily, by the spreading of vegetation and the changed
conditions which this causes, they have permitted ex-
tensive transportation and migration of animals. The
specific gravity of wood, its buoyancy in water, and the
various sails, vanes, and structures which favor wind
dispersal, and all the hooks and claws which help make
various kinds of burrs, and the edible fruits which animals
devour, all combine to favor transportation by wind,
water, animals, and other active agencies. By these
means, animals living within, or on such transported
parts, may be buoyed and transported by waves, currents
of water or air, and be carried by animals for long ir
tances, and into new localities and conditions.

(b) Animal Migration.—The movements of animals
which take them from one locality to another are exceed-
ingly diverse. They vary not only with the character of
the animal, at different stages in its life history both
structurally and functionally, and also under varied en-
vironmental conditions. The fixed and sessile animals
more nearly approach the conditions found commonly
among plants, but among animals this habit is confined
mainly to aquatic animals, that is, to animals living in a
mobile medium, which transports them at some stage.

64 THE AMERICAN NATURALIST [Vou. LIII

The most important character which influences the mi-
gration of animals is its own powers of movement. These
movements are dependent upon the ecology and the
physiology of the animal, its structure and its mode of
response. The general principles of response have
been discussed in the first part of this paper, where
the systems of activity, the cycles of activity, the
limits of activity, and the interaction of all systems
was emphasized. All of these factors should now
be recalled. Animals creep, walk, swim, and fly, accord-
ing to the media in which they live, their structure, and
their ecology, and the interaction of all these factors put
limitations upon animal movements.

From the standpoint of function, animal movements
and migrations have two main influences. By move-
ment the animal subjects itself to new conditions, these
conditions have a direct influence upon the animal, and
change the direction or its internal changes, and it be-
comes acclimated or dies; or by its repeated responses
and retreating movements, it escapes from the adverse
conditions and finally comes to rest in a new relatively
favorable condition (Adams, 715, p. 12). This monoton-
ous cycle is repeated with all the variations which diver-
sity of animals and diversity of conditions can produce,
and in its essential features it is the same from Protozoa
to man.

The geological age in which we live is one in which the
land surface, relative to the sea, has accumulated uplift
from former ages, and has been newly elevated, and as a
result there are many high mountains, and the seas are
relatively deep. These are conditions of stress, and the
processes of adjustment to strain are in full operation.
This is a period of relative diversity of the lands and
of the seas, which favors diversity, both in the atmos-
phere and in the hydrosphere. With the elevation of
the land, this diversity is shown both vertically and
horizontally. Large areas lie at considerable altitudes
and in their departure from the narrow mean surface

No. 624] MIGRATION A FACTOR IN EVOLUTION 65

stratum, are subjecting many animals to stress, and to
the process of acclimatization to the high altitude con-
ditions. The climatic diversity shown vertically, also ex-
presses itself longitudinally, by interference with free
circulation of temperature, moisture, and other climatic
factors, and tends to produce the varied climatic zones,
such as tropical, temperate, and polar, as well as the
diversity due to humidity. These diversities mean that
many minor circulating systems are caused and conse-
quently there are varied local wind systems, further
favoring diversity. All of these influences tend to favor
local or relatively limited migration, rather than the
widespread dispersal of animals subjected to such con-
ditions.

The hydrosphere is influenced similarly. The diversity
of the lands favors diversity in inland waters, and the
arid climates favor saline waters. Inland waters under
these conditions tend to be isolated and varied. In the
sea the deepened waters produce stresses similar to
those on land produced by altitude, and the elevations
and relative increase in the land area interferes with the
circulation of marine waters and favors local differences
and local stresses. With the deepening of the seas, the
shallow waters are restricted and the littoral animals are
under increased stress. It is seen, therefore, that the
conditions dominant on a world seale are those of stress
or strain, now in the process of adjustment. It should
be observed that all spheres, the atmosphere, hydro-
sphere, lithosphere, and the biosphere are involved in the
same general interacting process. It is only by viewing
the subject broadly that we thus gain this perspective of
the status of our own times.

We may now turn to certain details which will help to
illustrate the application of these ideas to animal migra-
tion. The same grouping of influences will be used which
has been applied in the preceding discussion of passive
migration.

(a) Atmospheric Influences.—The atmospheric factors

66 THE AMERICAN NATURALIST [Vote LIHI

which influence active animal migrations are mainly
those which are dependent upon: chemical composition
(largely oxygen, CO, and volcanic gases); temperature;
pressure; humidity; and mechanical effects, as they are
combined in climatic.changes. All of these influences un-
dergo differences which influence, by acceleration or in-
terference, the movements and migrations, acclimatiza-
tion and ecological attunement of animals. Furthermore,
these influences, or their systems, do not act independ-
ently but at the same time, so that their laws of inter-
action are the main rules of the game.

(b) Hydrospheric Influence.—The hydrospherie influ-
ences are similar to those of the air, depending on: com-
position (salts and gases); temperature; pressure; and
mechanical effects (waves, tides, circulating currents).
In this geological age of stress and diversity, on account
of the mobility of this medium, it has transmitted its
pressure with slight friction to animals. Upon land the
_ active streams are a direct response to the steep slopes
down which they flow, and they visualize at once the
reality of this active media which has kept fishes and
other animals busy moving up stream for millions of
years. Bodies of standing water, by their isolation, ex-
cept when connected with streams, tend to retard active
animal migration. Such bodies are likely to abound in
the early stages of uplift and to decline as drainage lines
develop. The diversity of climate into humid and arid
as previously mentioned, tends to favor diversity, chem-
ically, in bodies of standing water. These inland waters,
while discontinuous to many aquatic animals, are not so
to many flying and running kinds.

In the sea, the narrowing of the continental shelf tindi
to crowd the shore animals, and favors isolation and di-
versity of habitat, and retards ready migration for many
animals, The hastened erosion, however, tends to in-
crease the continental shelf and its continuity. The
deeper water shows relative diversity and tends toward
stagnation in its currents because of the relative increase
of the land area.

No.624] MIGRATION A FACTOR IN EVOLUTION 67

(c) Lithospheric Influences.—The interdependence of
the physical spheres is so marked that by this time, in the
discussion of the air and water, the solid earth has been
included in part. The rigidity of the lithosphere is so
great that its adjustments to strain are in general rela-
tively slow and of long duration. The density of the
medium is so great that animals inhabit only a shallow
surface stratum, the upper part of the zone of weather-
ing processes, in or on the soil. Animals living in the
soil are influenced by its density, its physical and chem-
ical composition, temperature, and its movements. Those
living on it are relatively independent of the quali-
ties just mentioned, but are much influenced by the relief
of the surface, by the climate and vegetation, and are
more truly air rather than earth dwellers. The elevation
of the land above sea in itself, and not as modified by
climate and topography, probably has little direct influ-
ence, except in its degree of stability with regard to
erosion. The greater the altitude and the steeper the
slope, the greater the physical stress and the rapidity of
erosion. Ice and landslides exert pressure and drive
animals before them, and are largely dependent on eleva-
tion and slope. The cycle of degradation of the land,
particularly its topographic diversity, greatly influences
the degree of freedom in the movement of land animals. .

(d) Plant Influences.—The stresses of the physical en-
vironment in the air, water and earth, impose pressure
upon the vegetation. Since the largest number of ani-
mals are directly dependent, and a smaller number indi-
rectly upon plants, much of this pressure is transmitted
to animals. The climatic diversity, seasonal and secular
cycles, influence the amount of animal food. Some ani-
mals, during adverse seasonal conditions and scanty food
supply pass into an inactive state, and tide over such a
season, and most animals not possessing such tend to mi-
grate. Thus upon the plains the bison wandered with the
seasonal changes of pasture, just as mountain sheep and
goats migrate up and down the slopes as their pasture

68 THE AMERICAN NATURALIST (Vou. LIL

varies with the season. The succession of vegetation
upon all surfaces, drives out some animals, just as it in-
vites others to follow with it, as when, with the develop-
ment of forest, the animals of the open find conditions
unfavorable. The kind of vegetation, whether herba-
ceous, woody, conifer or deciduous, ete., has an impor-
tant influence upon the movements of many animals.
The cycles of vegetation also change the physical condi-
tions, the temperature, humidity, soil conditions, and
thus influence animals.

(e) Animal Migrations.—The individual migrations of
animals, caused by their own spontaneity, and that by

> a So
[Areonericat Piai Deia JA | : 3
THE UPPER i

fl DRAINAGE AREA

ir
Fic. 1. Map showing the hypcthetical Permian drainage of the Upper Ten-
nessee drainage area. Compare with the present drainage shown in Fig. 2.

other animals, are exceedingly varied. All the factors
which influence individual movements, as indicated in the
first part of this paper, now apply in detail, and in addi-
tion there is the pressure exerted by animals living asso-
ciated with them. Simple animals require a favorable

No. 624] MIGRATION A FACTOR IN EVOLUTION 69

environment as truly as complex ones. It is known that
many animals decline in vigor if kept in the same medium,
but if the medium is kept fresh, or the animal moves
about freely and secures a fresh medium, it thrives.
Freedom of movement thus permits the animal to move
away from influences which interfere with its system, and
thus minimizes the disturbance. The continuation of
this process tends, with freedom, to bring the animal ul-
timately into favorable non-interfering conditions if

oS oF gs

PRESENT DRAINAGE

OF THE UPPER
TENNESSEE RIVER SYSTEM _,J

such exist. With these ideas in mind we need to recall

Į

that the atmospheric, hydrospheric, lithospheric and

vegetational pressure all combine to encroach upon the
animal, and to interefere or reinforce its activities and
movements. The normal movements of the animal, and
the ordinary routine of environmental changes, are thus
in process of continual adjustment. Thus with the mi-
gration of the animal habitat, whether caused by a change
in the atmosphere, hydrosphere, lithosphere, vegetation,

70 THE AMERICAN NATURALIST [Vou LHI
or any combination of these, the animals also tend to
migrate with it and they are thus led about all over the
surface of the earth. In this we see the importance of
cycles of climatic changes, cycles of crustal movements
of the earth, changes in circulation of the sea, and the suc-
cession of plant and animal associations. It is to the
changes of this character that the student of paleontolog-

a
gi Oe
EA gr “oy
- -3 meene iaa yl ins A e acme Siem woo
maemo j A, “ Re i? TA y Bucs. Pe
a <
WR ae
Æ mA ii
YE K tt Jae a
Oo Ke 5 ’ A 7
> PA N A 3 . K
4 N o we a oe
ne phe Zs ueadvace KS : oe
Fe È eor ta
i D A ` {
[is Louoow
| i
far
Bi Vi
y
Vir A
id
pf
Lys
J ~
| Wenara

Fie Map showing the hypothetical migrations of the snails of the genus
Io in the upper Tennessee River system, as influenced by drainage changes. The
numbers refer to the kind of shell illustrated in Fig. 5.

ical evidences and causes of evolution gives much thought,
and it is to the present evidences of these changes to
which the field ecologist gives much attention.

In my study of migrations of the fresh-water snail Jo,
in the drainage of the upper Tennessee River system
(Adams, ’15b), it was found that there were great cycles
of change in the history of the streams, and that there
were probably corresponding migrations of the snails.
This is shown if we compare the map of ancient hypo-
thetical drainage, Fig. 1, with that of the present, Fig. 2,
and the supposed migration of the snails, Fig. 3, and
compare these with the map of their present distribution,

No. 624] MIGRATION A FACTOR IN EVOLUTION 71

Fig. 4. The shells of these snails are shown in Fig. 5.
The presence of these snails in the headwaters of streams
appears to be due to the ordinary creeping movements of
the snails taken in connection with the up-stream migra-
tion or growth of the stream habitat, because, on the
other hand, the current tends only to favor a down-stream
dispersal. Such animals, therefore, appear to be led
about by the migration of their habitat. This sort of
migration is comparable to those land migrations which

Fra

Fic. 4. Map showing the present distribution of the forms of the snail Io
in the upper Tennessee River system. The numbers refer to the kind of shell
illustrated in Fig. 5.

have clearly taken place during climatic migrations, as
during the ice age, and during similar changes in humid-
ity, and with base-leveling ee (Woodworth, ’94;
Adams, ’01).

The competition among different kinds of animals has
long been recognized as an important factor in animal
migrations. Overcrowding produces a condition of
stress, and as a result of this stimulus, animals tend to
migrate and become diffused from the region of pressure
in all possible directions. Thus new conditions are en-
countered which necessitate changes on the part of the
animal, and thus this process continues indefinitely.

ou. LIL

j

[V

NATURALIST

'AN

THE AMERIC

2 SL UL poyRoIpU re suopeaigpu [LOTJeyJOdAY əsoqm ‘OF snuəg oy} UT SPYS JO swIogz ueu Əy} JO SUOPPLIJSNIJJI Q 'ÐIA

as

No.624] MIGRATION A FACTOR IN EVOLUTION 73

IV. Summary AND CONCLUSIONS

The animal should be looked upon as a dynamic system
which tends to continue in its course of action until
changed from within or until diverted by external inter-
ference with its system, and until a condition of relative
equilibrium is developed by balancing all influences. The
behavior of animals should be viewed as a process of
rhythmical activity.

The cycle of activity of the animal agent is a unit of
_ fundamental importance. To study cycles, their dynamic

status, their degree of relative equilibrium must be de-
termined. In this manner the conditions of stress, the
processes of adjustment to strain, and the conditions of
relative equilibrium may be recognized and determined.
These determinations should be applied to all cycles of
activity, that of the life history, and all others. The use
of these ideas enables one to apply Bancroft’s law—that
a system tends to change to minimize external disturb-
ance—to animal activities, and thus one is enabled to ex-
plain a large number of diverse observations. Supple-
mentary to Bancroft’s law are the influences which tend
to accelerate or reinforce, without other change, the con-
dition of the animal.

The activities of animals cause them to collide with
their environment. Conditions under which animals
have become accustomed or attuned are those of relative
equilibrium. With departure from these conditions, the
animals are stimulated, their system is interfered with,
and the animal tends to change until the interference is
minimized. The hindrance thus placed upon animal
activities are its ‘‘limiting factors,’’ and these are to be
viewed according to Bancroft’s law. This law is not
limited to the actions of the individual animal, but in-
cludes also the race, and those of animal associations.
The Vernon-DeVries law of the diminishing influence of
the environment progressively during ontogeny, is an
example of limiting factors according to Bancroft’s law.
This law of Vernon’s is of great value in the study of

74 THE AZIERICAN NATURALIST [Vou. LIII

migration in relation to evolution because it suggests the
critical period at which the stress of the new environ-
ment may have its greatest direct influence upon the new
generation and thus influence its heredity.

The next important category above the animal system
is the law of interacting systems. The main models of
interacting systems are: .

1. The physical model of interacting forces, recalling in
this connection the law of inertia, the tendency of a body
to continue in its present state at rest, or in motion, and
the law of reinforcement or acceleration.

2. Bancroft’s law is that a system tends to change in
such manner as to minimize external disturbance. This
should be applied to the interaction of all systems. This
is a law concerned with responses to stress and to the
process of adjustment, and it shows development or evo-
lution of equilibria.

3. The phase rule applies to the result of responding
to stress or equilibria. This is thus complementary to
Bancroft’s law; one is concerned with the condition of
stress, and the other with the condition of equilibria.

These laws appear to be universal and not limited
solely to the non-living. Irritability may not be causally
explained, but it seems to obey these general laws in the
same manner as causal changes. Applying these laws to
animal migration, we see that the present geological age
is one of physical stress, and that the process of adjust-
ment to strain is now in operation. The physical stress
applies to the air, water, earth and to their interactions.
‘This is an age of physical diversity—tending toward one
of simplicity and uniformity. With diversity there are
many local cycles of activity in all features of the en-
vironment. These cycles of circulation influence the
transportation of animals, and their active migrations.
By transportation and migration animals encounter new
conditions, new stresses, and change to minimize the dis-
turbance and acclimate themselves to the limit of their

No. 624] MIGRATION A FACTOR IN EVOLUTION 75

possibilities; and they repeat this cycle with unending
monotony and persistence.

December 15, 1917.
V. BIBLIOGRAPHY
Abbe, C.
1908. The Progress of Science as Illustrated by the Development of
Meteorology. Smithsonian Report for 1907, pp. 287-309.
Adams, C. C.
1901. Base-leveling and its (ore: Sie wera: with Illustrations from
utheastern United Sta i ; . 839-85
1904. On the Analogy AE te Departure from Optimum Vital
Arog and ay sips aad Geographie Life Centers.
Science, N. S., Vol. 19, pp. 2
1908. The Took pteail Succession i bua The Auk, Vol. 25, pp.
109-153.

1909. Isle Royale as a as Environment. Ann. Rep. Mich. Geol.
Surv. for 1908, pp. F
1913. Guide to the Study of pate Ecology. Pp. 1-183. New York.
1915. An Outline iat the Relations of Animals to their OnE Environ-
ments. . Ill. St. Lab. Nat. Hist., Vol. 11,
1915a. An Ioann Study of ing me Toiset Sete sa Bull.
Il. St. Lab. Nat. Hist., Vol.
19156. The Variations and tisotogtea? Shisha of the Snails of the
Genus Io. Memoirs Nat. Acad. Sci., Vol. 12, Pt. 2, pp. 1-184.
Bagehot, W.
1873. ef ne and Politics; or, Thoughts on the Application of the
Principles of ‘‘Natural Selection’’ and ‘‘Inheritance’’ to
Political Society. Pp. 1-224. New York.
Bancroft, W. D.
A Universal Law. Science, N. S., Vol. 33, pp. 159-179; hes
Jour, Amer. Chem. Soc., Vol. 33, pp. 92-120, 191
Blackman, F. F.
1905. ans and Limiting Factors. Ann. of Bot., Vol. 19, pp. 281-

1908. ie Manifestations of Chemical Mechanics in the he Plant.
ritish , 1908, pp. 1-18 (separate).
Blackman, F, F., and Smith, A. M.
1 Experimental Researches on Vegetable Assimilation and Respira
IX., On Assimilation in Submerged Water-Plants, Di
ke Relation to the Concentration of Carbon Dioxide and other
Factors. Proc. Royal Soc., B, Vol. 83, pp. 389—412, 1910.
Brooks, W. K.
1902. The Intellectual Conditions for Embryological Science. Science,
N. S., Vol. 15, pp. 481-492,
Chamberlin, T. C.
1906. On a Possible Reversal of Deep-Sea Circulation and Its Influ-
ence on Geologie Climates. Proc. Am. Phil. Soc., Vol. 45, pp.
1-11.

76 THE AMERICAN NATURALIST [Vor. LIII

wee T. C., and Salisbury, R. D.
4-06. Guotory. Vols. 1-3. New York,
ona 6. M.
1915. A Dynamic Conception of the reggie Individual. Proc. Nat.
daa tik pp. 164-172

915a. Senescence and Rejuvenescence. Pp. 1-481. Chicago.
rag Individuality 3 in Organisms. Pp. 1-213. Chicago.
Clements, F. C.
1916. Plant Succession, an Analysis of the Development of Vegetation.
Carnegie Inst. of Wash. Pub. No. 242, pp. 1-512.
Conant, C. A.
1908, The Influence of Friction in Economics. Science, N. S., Vol. 27,
pp. 99-104
Cowles, H. C.
1911. T one of Vegetative Cycles. Bot. Gaz., Vol. 51, pp. 161-

Davis, W. M. Gali ted by D. W. Johnson.)
1909. Geographical Essays. Pp. 1-777. New York.
Fassig, O.
1899. Types of March Weather in the United States. Amer. Jour.
Sci., (4), Vol. 8, pp. 319-340. (This paper, taken in conjunc-
tion with others listed in this bibliography, will materially as-
sist the zoologist in the applications of the process conception
to the atmospheric problems with which he deals. Cf. Abbe, 708.)
Findlay, A.
4. The Phase Rule and its Applications. Pp. 1-313. London.
Henderson, L. J.
1 The Fitness of the Environment, An Inquiry zt the Biological
Significance of Matter. Pp. 1-317. New Yor
1917. bes Order of Nature. An Essay. Pp. 1-234. Clan Mass.
Hooker, Jr.,
917. ae s Law of the Minimum in Relation to General Biological
Problems. Science, N. S., Vol. 46, pp. 197-204
Huntington, B.
1915, Civilization and Climate. Pp. 1-333. New Haven, Conn.
Jennings, H.
1906. Retavior of the Lower Organisms. Pp. 1-366. New York.
1912, Age, Death and Conjugation in the Light of Work on Lower
Organisms. Pop. Sci. Mo., Vol. 80, pp. 563-577.
1913. The Effect of a aa in Paramecium. Jour. Exp. Zool.,
Vol. 14, pp. 279-391
Keyes, C. R.
1898. The Genetie Classification of Geological Phenomena. Jour.
Geol., Vol. 6, pp. 809-815.
pair sao ù
1917. À Quarter-Century of Growth in Plant Physiology. The Plant
World, Vol. 20, pp. 1-15.
Mellor, J. W.
1904. Chemical Statics and Dynamics, including the Theories of Chem-
ical Change, Catalysis and Explosions. Pp. 1-528. London.

No.624] MIGRATION A FACTOR IN EVOLUTION 77

Pierce, W. D.
T916. A New Interpretation of the Relationships of Temperature and
Humidity to Insect Development. Jour, Agr. Research, Vol.
5, pp. 1183-1191.
Pike, F. H.
1917. The ond of Death. Jour. of Heredity, Vol. 8, pp. 195-199.
Pike, F. H., Soa Seo
1915. The Peon ‘of Certain Internal Conditions of the ee
in Organic Evolution. AMER. Nat., Vol. 49, pp. 321
TS , A.
1910. The Influence of Darwin on the Study of Animal Embryology.
Pp. 171-184. Darwin and Modern Science. Cambridge.
Shelford, V. E.
1911. Physiological Animal Geography. Jour. of Morph., Vol. 22, pp.
551-618.

Thompson, D. W.
1917. On Growth and Form. Pp. 1-793. Cambridge.
Thorndike, E. L
1911. Niiet Intelligence. Pp. 1-297. New York.
Van Hise, C. R.
1904. The Problems of Geology. Jour. Geol., Vol. 12, pp. 589-616.
Vernon, H. M
3. Vatiation in Animals and Plants. Pp. 1-415. London.
Willis, B.
1911. What is Terra Firma?—A Review of Current Research in
Isostacy. Smithsonian Report for 1910, pp. 391-406.
Woods, F. A.
1910. Laws of pnyin Environmental Influence. Pop. Sci. Mo.,
, pp. 813-
Woodworth, J. B.
1894. The Relation between tengeh and Organice Evolution.
Amer. Geologist, Vol. 14, pp. 209-235

To one interested in some of the broader Dooa of dynamic ideas,
the following additional papers will prove suggestiv:

Hale, G. E.
1912-13. eed Other Worlds. The World’s Work, Vol. 25, pp.
66-182 (1912), 286-302 (1913). Special attention is called
to the second part of this paper on the relation of gravitation
to radiation as a cycle of transformation.
Herrick, C. L.
1906. pipat of Dynamic Theory to Physiological Problems.
J Comp. Neurol. and Psych., Vol. 16, pp. 362-375
Keyes, C. R.
1912. Deflatative Scheme of the Geographic Cycle in an Arid Climate.
Bull. Geol. Soc. Amer., Vol. 23, pp. 537-562.
Spencer, Herbert
1913-15. The Principles of w Vol. I, pp. 1-706 (1898); Vol.
II, pp. 1-663 (1899), N. Y

78

THE AMERICAN NATURALIST [Vou. LII

1916. First ie pp. 1-550. New York. Sixth edition of 1900.
Althoug

Very, F. W.

hr these books by Spencer are in many ways not

biological problems which uses so many. ae cae dy-
namic conceptions. After familiarity with more modern views,
these books may be read with great profit, particularly his dis-

win and Modern Science, 1910, pp. 450-455.

1902. A Cosmic Cycle. Amer. Jour. Sci., (4), Vol. 13, pp. 47-58; 97-
114;

185-196.
1913. Sas becomes of the Light of the Stars? Pop. Sci. Mo., Vol.
89-306.

2, pp.

ERRATA

P. 471, line 23 from top, third word should read o”
P. 472, line 5, for ‘‘ontology,’’ read ‘‘ontogeny.’’
P. 482, line 29, for ‘‘really,’’ read tf readily.’’

SOME STUDIES IN BLOSSOM COLOR INHERI-
TANCE IN TOBACCO, WITH SPECIAL REFER-
- ENCE TO N. SYLVESTRIS AND N. TABACUM

H. A. ALLARD

U. S. DEPARTMENT OF AGRICULTURE

Tae blossoms of varieties of Nicotiana ee exhibit
three distinct colors, white, carmine and pink.

In the writer’s crossing experiments, two vite.
ered nicotianas were used, N. sylvestris, a species with
long, slender, pure white blossoms, and a variety of N.
tabacum from Honduras (S. P. I. No. 30887), with rather
small, pure white blossoms of the tabacum type. The
pink-flowered variety generally used was the Connecticut
Broadleaf variety, although the varieties 70-leaf Cuban,
a mammoth type of Cuban which appeared as a mutation
in Connecticut in 1912,and Maryland Mammoth also were
used. The carmine-blossomed tobacco? is a variety of
tabacum sold by various seedsmen for ornamental pur-
poses under the name giant red-flowering tobacco. This
variety breeds true to blossom color and crosses readily
with all the commercial varieties of tabacum.

CROSSES OF PINK-FLOWERED VARIETIES WITH CARMINE-
FLOWERED VARIETIES

In the crosses Pink 9 X Carmine ¢ and their recip-
rocals, Carmine, without exception, has been perfectly

1 The colors carmine and pink have been compared with Ridgway’s Color
Standards and Color Nomenclature, 1912 edition. The carmine is practi-
cally identical with Ridgway ’s carmine, shown on y 1. The pink matches
almost exactly his Hellebore Red, shown on Plate

2 There seems to be little definite information à hand soneerning the
origin of the carmine-flowered varieties of N. tabacum. O. Comes, in his
monograph ‘‘ Delle Razze Dei Tabacechi,’’ Atti. Del R’Inst.d araa teekito
di Napoli, Serie 6, 1905, pp. 77-306, speaks of the Nepal tobacco, a variety
of N. tabacum, as having intensely red blossoms. The Rano variety is also
said to have distinctly red blossoms,

79

80 THE AMERICAN NATURALIST [Von LIIT

dominant, so that all the blossoms of first generation
plants bear carmine flowers.

TABLE I
FIRST GENERATION PLANTS OF CROSS CARMINE X PINK
Year | Row Cross poe Remarks
1915 | 26A | Carmine 2 x Pink (Md. Mexpenot) Pi. ae 21 | Allcarmine
1916 | 21C | Pink (70 leaf Gaan G- X Mine esena: r ae is af
1916 | 26A | Carmine 9 X Pink (70 leaf ETS RST OTA m
1916 | 35A | Pink (Conn ea f) 9 X Carmine g. pnr z
1917 | 114A | Pink (Conn. Broadleaf) 9 X Carmine g. 50S EN
Tot eA a in lee AN EASES Ea 9s ' 146 |All carmine
TABLE II
SECOND GENERATION PLANTS OF CROSS CARMINE X PINK
Year | Row Cross Diente | Red | Pink
1915 | 27B | Carmine 9 X Pink (Md. Mammoth) g' 15 13 2
1916 | 19B | Pink (Conn. Broadleaf) 9 X Carmine à 39 30 9
19 ink (Conn. Broadleaf) 9 X Carmi 1 6 4
1917 | 91B | Pink (Conn. Broadleaf) 9 X Carmine g 42 29 13
1917 | 107C | Pink (Conn. Broadleaf) 2 X Carmi 9 4
1917 Pink (Conn. Broadleaf) 9 X Carmine F. 153 114 | 39
OGM re AS ED EEEN EE E ee eee 282 21i Ti

From the data shown in Tables I and II, it is evident
that the characters pink and carmine behave as typical
unit characters, with carmine completely dominant.
Segregation takes place in the 2d generation into carmine
and pink blossomed plants very close to the theoretical
ratio of 3 to 1.

A heterozygous plant of the first generation of the
cross Pink (Conn. Broadleaf) 9 Carmine ¢ was then
crossed with homozygous carmine. Of 115 plants ob-
tained in this cross, all were carmine in color, which is
in accord with the expected result.

Heterozygous plants of the first generation were now
crossed with recessive pink with the following results.

From these results it is evident that the theoretical
ratio 1:1 which obtains in such a cross is very closely ap-
proached.

No. 624] BLOSSOM COLOR INHERITANCE IN TOBACCO 81

TABLE III

CROSSES BETWEEN HETEROZYGOUS PLANTS OF THE First GENERATION OF
THE CROSS (PINK 9 X CARMINE ĝ) AND PINK

Year. | Row. Cross. sah Red | Pink
1915 | 18A | A first generation plant of the cross [Pink
(Conn. Broadleaf) 9 X Carmine g] 9 X
Pink (White Burley Mammoth) g ..... 19 T 12
1915 | 22 A first generation plant of the crozs [Pink
(Conn. Broadleaf) 2 X Carmine o’] 9
ink ite Burley Mammoth) ~..... 41 17 24
1915 | 26B | A first generation plant of the cross [Pink
erg aig cpio 2 X Carmine d] 2 X
Pink (Md. Mammoth) o.............. 19 11 8
1915 | 28B ae generation "plane of the cross [Pink
(Conn. Broadleaf Carmi
Pink (White Burley Mammoth) g ..... 16 7 9
1917 first generation plant of the cross [Pink
(Conn. Broadleaf) C ;
Pink (White Burley Mammoth) g ..... 197 91 106
1917 A first generation plant of An cross [Pink
C Broadleaf) 2 X Carmine g] 2? X
ink (Conn. Broadleaf): Scorers 249 130 ' 119
TP E a a cos es Va ae oo 541 | 263 | 278

In further studies of the cross Carmine X Pink, the
progenies of many extracted, recessive pink-flowered
plants of the second generation have been grown, and
all have produced pink-flowered lines. Of the carmine-
flowered plants of the second generation, some (the
heterozygous) have again broken up into Carmine and

ink, while others (homozygous) have given pure Car-
mine lines.

Crosses Invotvine CARMINE oF PINK witH WHITE

In crosses involving Carmine or Pink with White,
white has behaved as a recessive, whether the cross has
been made with the white-flowered species N. sylvestris,
or with the white-flowered variety of N. tabacum S. P. I.
No. 30887 from Honduras. Carmine, however, is not
perfectly dominant in these crosses for plants of the
first generation are somewhat lighter red than the car-
mine itself. In the cross Pink X White (N. sylvestris)
the blossoms of the first generation are somewhat paler
than the usual pink of the tabacum varieties. In crosses

82 THE AMERICAN NATURALIST [Von. LIII

between pink-flowered varieties of N. tabacum and
White (S. P. I. No. 30887 from Honduras) white has be-
haved as a recessive. The intensity of the dominant pink
depends upon the pink variety used. In some crosses,
the blossoms of the first generation plants are similar in
color to the pink of the pink parent. In other crosses,
the pink of the first generation plants may be noticeably
deeper than that of the pink-flowered parent.

A number of heterozygous plants of the first genera-
tion of the cross Pink (Conn. Broadleaf) 9 X Carmine g
have been used as mother plants and crossed with the re-
cessive white N. sylvestris, with the following results:

42 plants, of which 25 were carmine blossomed and 17
pink blossomed.

41 plants, of which 23 were carmine blossomed and 18
pink blossomed.

Here we have a total of 83 plants, of which 48 possessed
Carmine blossoms and 35 possessed Pink blossoms, show-
ing an approximation to the 1:1 ratio. In crosses in-
volving the species N. sylvestris, some difficulties are in-
volved, since the first generation plants are usually sterile
or nearly so. However, this sterility has been overcome
in the cross in which a first generation plant of the cross
(Pink (Conn. Broadleaf) 2 X Carmine ¢) was pollinated
with the pollen of N. sylvestris. In the second generation
of this cross, whites, pinks and carmines appeared. A
number of carmine plants were selected and their prog-
enies studied. One known as no. 12, proved to be homo-
zygous for carmine and has bred true to this color for
several generations. A sister plant no. 9 with carmine-
blossoms proved to be heterozygous. In a progeny of 32
plants obtained from this plant, 26 were carmine and 6
were pink blossomed, approximating the theoretical ratio
hep |

In the cross Carmine X White, using the white-flowered
variety of N. tabacum S. P. I. No. 30887 from Honduras,
the plants produce an abundance of fertile seed. As has
been stated, carmine is dominant over white, but it is not

No. 624] BLOSSOM COLOR INHERITANCE IN TOBACCO 88

a perfect dominance as in the cross Carmine X Pink, for
the blossoms of the first generation plants are somewhat
paler than pure carmine. In the second generation there
is a segregation into whites, and various degrees of pinks
and reds, ranging from light pink to pure carmine. Of
278 second generation plants of this cross, grown in 1917,
the blossom colors were grouped as follows:

BOLO oss ey P Ga ee Pua se eee he ee cans 65
Dark pibk sy: Hie sie Motes CAS AA aba i 26
Laight pink iaat hg. Ces SL aN Ce es 38
Liphtor TOAD COIOMNOS so aaa <b de ake decane aes 95
APUG oss ae enc cas ee ee be ee ae ta ea ec aes 54

GURL SEs Sa ie Re See eek Crees 278

It is evident that the recessive whites which were easily
determined approximated very closely the theoretical 25
per cent. Owing, however, to the uncertainty of analyz-
ing correctly the varied pinks and reds, the carmines
which probably represent the homozygous dominants
are somewhat below the theoretical 25 per cent. It is
quite probable that this class is somewhat smaller than
it should be, owing to the difficulty of distinguishing with
certainty all the homozygous carmines trom the hetero-
zygous somewhat paler carmines.

A number of extracted, recessive whites of the second
generation of this cross have been grown and all have
produced white-blossomed progenies. These white-blos-
somed plants have proved somewhat puzzling, however,
for some seemed to reveal a hint of color, like an almost
imperceptible sheen, that could be detected only in cer-
tain lights. So fugacious was this suggestion of color,
that I felt inclined to ascribe it to the imagination, until
certain crosses were made with pink-flowered varieties
as follows:

From the cross Pink (Maryland Mammoth) 9 X White
(extracted) g, 54 first generation plants were obtained, of
which 17 were carmine, 18 were pink as in the Mammoth,
and 19 somewhat lighter than carmine.

In the reciprocal of this cross, 7. e., White (extracted)
9 X Pink (Md. Mammoth) g, 43 plants were obtained, of

84 THE AMERICAN NATURALIST [Vou. LIII

which 20 were carmine, and 23 were pink blossomed as
in the Mammoth.

This same extracted white-flowered plant was also
crossed with Pink (Conn. Broadleaf) as follows:

From the cross Pink (Conn. Broadleaf) 2? X White
(extracted) J, 51 first generation plants were obtained, of
which 12 were carmine or approaching it, and 39 were
pink approximating the pink of the Broadleaf parent.
It is apparent that the extracted white used in these
crosses has somehow retained the factor of Carmine
which reappeared in the cross with Pink.

From the results reported in this paper, the Mendelian .
behavior of the cross Carmine X Pink is particularly
striking, and for those who wish to demonstrate before
students interested in questions of heredity simple Men-
delian behavior in a monohybrid, this particular tobacco
cross is especially suitable. Not only is the technique
of crossing easy, but a single fertile capsule from a cross
will produce several thousand seed. Furthermore, to-
bacco plants may be readily grown to the flowering stage,
in five or six inch pots under greenhouse conditions.

SuMMARY

Among the varieties of tabacum, carmine and pink in
crosses behave as unit characters, carmine being domi-
nant. In the second generation perfect Mendelian segre-
gation follows, approximating very closely the theoret-
ical ratio of three carmines to one pink. The extracted
recessive pinks and the homozygous carmines breed
true. The heterozygous carmines again break up into
carmine and pink. If a heterozygous plant of the first
generation is crossed with a pure carmine, the progeny
will all produce carmine blossoms. If it is crossed with
a pink, carmines and pinks will appear in the progeny,
approximating the ratio 1 to 1.

In crosses involving carmine or pink with white, white
behaves as a recessive, appearing in the second genera-
tion.

SHORTER ARTICLES AND DISCUSSION

A BIOLOGICAL SURVEY OF DESCRIBED CERCARLZ
IN THE UNITED SLATER

AMONG the earlier American zoologists Joseph Leidy alone
was a student of cercariæ. From his time to very recent years
American cercariæ have received little attention. This may
have been due to the greater demands made by other groups of
animals, or possibly to the minute size of the larve and a failure
to appreciate the exact differences of their structure. It could
not have been due to a lack of knowledge of the presence of cer-
cariæ, for the European records were abundant and the classical
studies of Leuckart, Ercolani and Looss had demonstrated the
life-history relations of cercariæ and adult flukes. Moreover,
the large number of adult trematode records showed that the
larve must be fairly abundant.

Within the past few years a revival of study in this larval
group has revealed a large number of forms, so that now there
are some sixty named species. Only eight of these have dates
prior to 1914. The majority of described cercariw have been
worked over by Cort, Faust and O’Roke.

A study of the descriptions of earlier named species shows
them to be very general, so that they apply not to the species at
all but to larger groups, genera or perhaps even subfamilies.
For example, the record of a monostome with three eye-spots
instead of characterizing a species merely distinguishes the tri-
oculate from the binoculate group of species. A parallel is
found in the diplostomulum commonly known as Diplostomum
cuticula von Nordmann 1832, which has been recorded from a
variety of vertebrate hosts and from equally variable habitats.
There is great probability of the existence of several new spe-
cies concealed beneath these generalized data. Such cases illus-
trate the futility of generalized descriptions.

Cort emphasizes the value of the excretory system of the cer-
caria as a basis of description. The conservatism of the system
is urged as a basis on which fundamental group relationships
of the trematodes can be discovered. Advantage in using this

* Contributions from the Zoological Laboratory of the University of
Illinois, No. 113

85

[Vou. LIII

THE AMERICAN NATURALIST

86

ueullseds əBurg ESgt “UBD ‘OJUOIOL uMouyun ysoy ‘umntiuenbe uy joe DIGI ‘PIEM 27YH1IM “DQ *oz
O'TI e161 “Uuo ‘pilose (Avg) wnsrep vuopdunp |` “FIGL ‘HOQ Dunuohh.y 9 “FZ
é é L181 “Vd “erydjepeyryg Avg sapoja ppuwh'T |` ` LLST ‘Aptary vaproosy +9 "gg
AavoH é L181 ‘ “Ske ‘erydjeperyd Avg snaind SULUD |" LLST ‘Apro vaproosp +9 ‘Zg
é į “Y ‘oremepeqd «Pouw pue
‘oupnpda ‘escola qt. Aueduioo bg ee ee ee ee a SoSI ‘ApIo'T sbo 9 "IZ
og GIGI “UBH ‘POUIIMVT ABS S1004} SUQLOUDIA |7 LIGI ‘PJOY,O vpnvsvubow +9 ‘OZ
Ov GTGT “ULY ‘eynueyD ‘PIEH vibaqus vsfiyd |****** "tt 29091d ‘2161 ‘POYO
8190460 ‘Q OId JSNUA suadsanonsb ‘O “BT
¿ SOGI “L'N ‘TANDA. ABQ onubaa SiIsMgoIUoy |" tt FIGL ‘HOQ vunpoBau +9 ‘SI
VT SI6T “TI “teary UouesuBg Avg UNDANI DILIIOMNAA |" "tt PIGI ‘HOO venyobaw *9 +21
sIwrogsıq
uəwrəds a]sulg TOGT “O ‘Aeg-ur-mda u{mouyun 4soy ‘uogyuvjd ur eer teete ‘9161 “PIM. nyoyda20u0b.10B + *9T
O'F GIGI “UBH ‘OBALNO ” ” ” se “*LIGT ‘SAOY,O 24.409“ “eT
OV GTGT “ULY ‘PIUIIMV'T ” ” tae Be “"""" "SIGE ‘HOD vydoysmp +9 ‘pI
o’¢ SI6L “TI ‘OFV ” ” no eal eae ras FIGI ‘H00 vydoysmp ‘9 “ET
oP GTGT “ULY ‘soueIMe'T ” ” ” FIGI ‘HOO s2pqoyur '9 “ZT
MOT €16L “MI “Bueqiy ” ” " poe FIGI “HOD sugqnyus Dp “TT
VI S16 “UVY ‘eousIMET Aeg saom siqiounj_ |*** ““PIGI ‘30D syrqoyur `) “OT
"somonsrgdury
Aavoy ¿ LLST Td “srydpperyda Aeg snaind srquoung |* "LLST ‘Aple'y muvon) ‘9 ‘6
é OSST “Lda ‘erydjepelyd ” ” pce Sr PST 1 Mn Dpnovopphy ‘9 °g
é ¿i SPST “Bd ettepend (Avg) vydoysosajay vsfiyg |********* SPST ‘UBUOpP[eH vpnvvopvhy P +2
£8 LI6I “TIT ‘touto Hy (Auoyjuy) oyayynd sisnqoruop js ** st test ttt eet SIGT “SNe Dpunn +p *9
0°09 L161 “IT ‘ARAA ABQ DUAD vsfiyd |****** "s**t tt SIGT ‘Sneg Dsengos "DQ *¢
Se 9T6T “JUON ‘SHJBAI0O VIT Duncoid Dwu |***** s**t ests LIGI “neg vpronpad ‘9 *F
og OTOT -FTOIT PROPRAN “27 AUS vund vshyg |' LIGT “Sneg vpmonyad +p 'g
Ste 9T6T * TP ‘SITBAI00 vor] VUOLE DDUWAT |" **** "sss LIGT ‘SNBA sisuappuoy +p g
o's SIGI “TI ‘eueqiy Avg Dus vsfiyg eee FIGI ‘HOQ sisuaungun “9 'I
SOULOJSOUOTY
woNsaJUT '4U90 Jog NWA puv AITB00T 4s0H OUI NT

cae ee CALIN HL NI WIavOUIH aaaraosaq wo ayooaxyy

I IYL

87

SHORTER ARTICLES AND DISCUSSION

No. 624]

ETGT “TII ‘ostoryD

Aeg E m

9%
09 LIGI “TI ‘QNA kie ona re e nee ene E FIGI ‘HOQ vazfjos081 '9 “Eg
00S 2161-OT6T TT Se eee a ee E FIGI ‘HOO 031702082 "9 “ZG
O'STI PIGL “TI ‘vuvqiy ABS suapoasy squoumd TR E YIGI ‘HOO vapfijos081 *D “Tg
9'er 9161 “guon “B[NOSSTI A BrT punrrosd voun Ce ee ee ee LIGI ‘qsne g DIDUILO Bre. “oe
Se S161 “usA ‘UEYN BT vuypuUD vsfiyd |"""* "°° PIGI ‘HOD DID *D “6h
é 6061 “LN ‘TVU ABg DNULL sispgovuos) FIGI ‘HOQ i0? "Dp “gp
£8 SIGI “UUOQ ‘pro;zIB TT Avg wnsiwap puodin |tt FIGL ‘3409 mypunomda "9 “LP
é & ¿eda ‘erydppepryg Avg puao TH | itrtereeeteets eLPST ‘APIO stoyay *9 ‘OF
é 6 2 Bq ‘erydppepryd Avg stuqnjoqyy 142 LEST ‘APIO stoyey wnantvo1a9 ‘GP
é ieee “eryd perq, 4 Arg pour. uou (stugsnan)) muonony |r thc yes 8981 ‘ Ap] ppp. Dap. 40} '9 “PP
é ¿i ¿ed ‘erydppepryg ABQ nyni (sman) ojuopouy |" BSST ‘Apro vposhipin) "Dp “Eh
é 9T61-FI6I OIN ‘DAU supsnog ABg vuopyoun oshiyg jtir BIGI ‘H00 180pnop ‘9 ‘Zo
é 9I6I-FIGBI “YONA ‘PYET SUNO Aqiamog vinjnbun vyouiivwma poumhiyT "ts SIGI ‘HOD aypun? *D ‘TP
¿ 9TI6T-FIGT ‘ "Wor ‘SJET svpsnoqd 4 ta sa of SIGI ‘H00 syunydaja ‘9 “OF
FST GIGI “UBy ‘souDIMeT Avg soara} siqioumd |" LIGI ‘P49YW.O MAND *D ‘6g
0'9 GIGI “ULY ‘MoLAoye'T 7 j A ETARE ` LIGI ‘SJO, O DPNDIOUIYI *D “BE
0z GIGIT “ULY ‘PDUIIME'T ü x is ` LIGI ‘PAOU, O PLL ‘O “2g

0'9 LI61 “I ‘qrq H js Ee A a gles a. I E AS SUES A E 900010
SIGI ‘QsnB i LOUTUL 9 oid DUNUVUL * “98
0°&% LIGT ‘ pus ‘Woop ould ROG TRIAD GAN fo a ee SI6I “snug SOOD “QD GL
0'001 LI6I ‘ "Il ‘eueqiy) + ee en ee en eee SI6L “qsneiy spb “v "Fe
0'99 LI6T “MI ‘AQA ABS 81470404} siqioum |tt SIGT “Wsneg SVDD +p “gE
e'g QTGI “JUON ‘SITTBAIOD LBS DUWO sfiyg tities LIGI ‘SNB Duos *D Zg
Te OT6I “JUON “EMOSSIN VIJ DUOLd vuu] LIGI ‘sneq DUMIMD `O “TE
P61 9T6I < mop ‘TOATY JOOY IOMA ABS vuriib osfiyg ttit LIGT ‘ISNA vuypovib “9 “Og
į QIGI-ST6I “YAWN ‘PYLT SUBO 1918 M MquvHd spubos ey, E E ne FIGI ‘H00 YRYMOP "DN ‘6z
é 9OIGI-ST6T “Yoru ‘PALI sepsnog Avg vssaiddp sypubos vouw |’ FIGI ‘4100 mymop *D "RZ
£01 S161 “TI ‘OBLO Ang vrayas pouwhiy \********** ‘FIGI ‘4100 mymop *D *1z
— FEST “TEJO “4g OYeT UMOUAUN soy ‘Juruurms-əang | 19T6L ‘PIB AA SapLosoyoUD "9 °*9Z

souloysiqy
4s07x, OuIB NT

uoloasuy “Qua Jed

978A PUL AII[800T

panuyuog—T WIV

[Vou. LIII

THE AMERICAN NATURALIST

‘UNUIO}SIP OURS g
“UNUIOFSIP OURS W ATQISSOg z

'LT6T “TII “eueqiy ‘umrenbe ue ut 0917 ‘pwojsosopw +9 pus ‘ZTEI ‘oy oyerT
‘quiog avpoy “ds ewopedurey WoIg ‘1494040040 “9 2 saprosoyoun ‘9 0} poep ATasojo saroeds om4} poqriosep seq Ipm Oy} ATJWODeqT q

é ‘puy ‘SMEH L
è 8d ‘srydjeperyg

(Cuury) synubojs pauwhy
(Avg) wnidoosn}n2 1q7DH

' 96ST "2244o 4q ‘ds *9 ‘Tg
EE + t+ OBBT

yee Aq PURI Beso (apfij090.173,.,) "9 *08

EEZ 9TGT “JUON ‘SITTBAIOD LUS DUAD oshyd |****"****** LIGT ‘NBI srmtofypoqvy *D 6L
səurogso oH

O'I GIGI “ULY ‘souermerT Avg DUURB voshiyg "*LIGBI ‘940.0 suusofisnf "9 *
0°00T LIGI “TIT ‘eueqiy ABg srajoary siquoumgd | SIGT PISNE DwoseyjunoD "DQ *LL
0'001 LIGI “TIT ‘eueqiy EEEE EELE, SI6L ‘qsne iq DULOISOYJUDID Fe OL
£8 LIGI “IIT ‘A9910 ourg Avg munan vsa Tae ""*""ST6L ‘snug vppuaposiys "Dp QL
OSL 9T6T “JUON “BPNOSSIN JA Aug sazons aqung |********** “**JT6T ‘Qsneg Divus 9
T6P orot ome a a a a LIGT “9sneq DDU 9 “eZ
D2 QIGI “JUON ‘BMOosSSIN ‘HI Aeg purihb pshud “"*****2T6T "goner vrazfeq “9 ‘zL
e'e LI6T “MI ‘AQAA Ur eee ***" “PIGT "HOO sagor “9 “TL
MOT S161 “MI ‘vueqra Aug szajan} siquoumda : “"*" "PIG ‘HOO swo -9 “OL
9°9T SI6L “UUOQ ‘prose xT (ABQ) unsiwap vuojadun p) x "'SPITGI ‘HOQ Duqns ‘O *69
OTS E161 “TIT ‘osvorqD Avg poxayfas pauwli'yT oer * "STG ‘HOD wrayer “QD “89

Jood ur vary i OAM “3pug yog ¿ds vauwA’y |" * "OOGST ‘Aplo'T vunhignd - +19
o'o LIGI ‘BMOT ‘Joda J a be S "SI6I qe ysod..vpf svyaLossvy *9 *99
a7 4 GIGI “UEY ‘PUIA ” n T eae "LIGI ‘HOY, O sisuaisuny * *g9
Lz GTGT “ULY ‘Spesso AVY 81070021) 819QL0UD] J ""'"2T6T ‘840.0 DDDB “9 “49
o'g SIGT “UVY ‘UIME PRR TS “" tet t ATOL ‘SHOHO 2480Y “2D “29
9'9 LIGT “TI ‘F099 utd ” ” ” ` BTGT ‘Isneq Diafiys 9 “ZO
0's SI6L “TIT ‘PIOPPO ABg vurwib vsiyg |" "* “"""PIGL ‘HOQ vinydopway “9 “19
£01 QTGI “JUOW ‘EMOSSIN “407 3 j k í e "'"LIGL ‘SNBA vsowasons +9 *
£99 QT6T “JUOW ‘ByNosstpAT a 23 eee ro ee aN LIGT SNBA ruhspydoinwm +9 *
ete 9I6L “UOT, ‘SIHJBAIOO vog vunrod pouwh'T : LIGI “sneg vunydnrp *p '
0'OF OI6T ‘QUOT, 'SITBAIOQ Avg vuwhib vshyg : ‘ALIGI ‘sneg vsopmpunð Dp *
£01 9T6T WHN ‘IOAN J00% IINA » T w 17t T LTGT “9snegq Doyupuap +p *
PAH QTGI “JUON ‘EMOSSIN ‘HI VT Duixoid pouwhTy te “"LIGI “Shey DNPupuap +9 *

SOUO}SICT
uonoəzur 47u90 d so ouren

9B PUL AIVOT

ponuyuog—I ATA ViL

No. 624] SHORTER ARTICLES AND DISCUSSION 89

system lies in the fact that it can be studied entirely in the living
cercariæ. The writer has used this method with profit, but in
addition has worked out a method of staining the genital organs
in the preserved larve. This method can be utilized when the
worker has access only to preserved larve. While the excretory
system is indeed a conservative system, the genital system is
probably more conservative and less likely to change from cer-
earial to adult stage. It has been found to be remarkably simi-
lar in the large, yet variable in minor, details in groups of cer-
cariæ known to be related through other organs or systems.
The best description of a cercaria is probably that which includes
both the excretory system as worked out in the living animal and
the genital cell masses as WORE in carefully preserved and
stained material.

A mere superficial description of the worm is a distinct burden
on the literature. The cercaria should be carefully studied in
minute detail or not at all. It is the nicety of technic and care
in observation which have yielded the number of species now
known and bids fair to increase the number vastly in the next
few years. It is necessary, then, to urge the investigator in this
group to use the utmost care in his work, to describe the minute
parts of the organs, and to record the complete biological data
available that these records may be of use in life-history inves-
tigations.

In order to place the more important biological data of de-
scribed cercariz in the United States in a convenient form, a
table has been prepared to cover the groups, the authors and
dates of the naming of the species, the hosts, localities and dates
of collections and the per cent. of infection (see Table I). The
same data have been collated from the standpoint of the host
in Table II.

A study of the described species shows that the great bulk are
distome larve. Most of these fall into three groups, the stylet,
echinostome and forked-tailed cercarie. The former group bear
evidence of relationship to the Plagiorchiide; the echinostome
cerearie are known to be larval Echinostomide, and the forked-
tailed cercariæ are probably larval schistosomes. The life his-
tory of only one species in the group has been worked out with
certainty, that of Cercaria Lissorehis fairporti, with Planorbis
trivolvis as larval host, a chironomid larva as intermediate host,
and Ictiobus spp. as definitive hosts. Of the species recorded

90 THE AMERICAN NATURALIST [Vou. LI

TABLE II
RECORD OF MOLLUSK INFECTION WITH CERCARIZ
|
Host Specie phn tee A O anah aaar | a
Planorbis wianek Gees oa ces 14 0 3 11 0 7
Planorhis partit: soros 60s 3 1 0 2 0 il
ample has or ae ae ae 3 0 0 3 0 1
ih ie RE Es oe ees 1 0 0 1 0 1
8 2 0 6 0 1
pea tinoate siete A angulata.. 1 0 0 1 0 1
j COO es PENS. 1 0 0 1 0 1
RAINE :.« (ecma's eexss 1 0 0 0 1 1
ymnea stagnalis appressa 1 0 0 1 0 1
yymnea stagnalis perambla 1 0 0 1 0 1
UMNE TRA ei dk wa do whe 3 0 0 3 0 1
Physa a eS Dc DOR cue 19 3 0 15 1 5
Physa antinati ds ia Oe. 1 0 0 1 0 1
Physa heteroatropha ES eh eee 1 1 0 0 0 1
PUG GAMA. eL pee a cs ce 1 0 0 1 0 1
Physa ancillaria.. 1 0 0 1 0 : Ai

Goniobasis pulchella......... 1 1 0 0 0 1
Goniobasis virginica.......... 2 0 0 2 0 1
Pleurocerca elevatu’ 1 0 0 1 0 1
G CALOSCOPÄUM » o siepi oaae 1 0 0 0 1 1
He m... 1 0 0 1 0 1
Helix albolabris. 66.02 82.0205. 1 0 0 1 0 1
cataracta. are oe 1 0 0 1 0 1
Anodonta marginata......... 1 0 0 1 0 1
Free-swi Mg ohara 0s 3 0 1 2 0 3

Total No. ponared = records| 72 8 4 57 3

In two or more h 11 2 0 9 0

Not epeciews 5.6 6. ee a Sy 61 6 4 48 3

for the United States only one, C. (Tetracotyle) typica Diesing,
1858, is recorded for another locality than North America. `

The larval hosts are without exception mollusks. All except
two, Anodonta cataracta and A. marginata, are Gasteropoda.
Several of the species have been found in two snails, although
none have been recorded as infecting three or more hosts.
Usually where the species occurs in two hosts the infection of
the one is more widely spread and heavier than that of the other.
Several records show the parasitism of several species of cer-
cariæ within the same host species in the same locality. In fact,
the writer found as many as four trematode species within the
same host individual (Planorbis trivolvis) at DeKalb, Illinois,
in August, 1917. The occurrence of two cercari species in the
same host individual is commonly found in the records. In this
ease one of the parasites usually has a heavier hold on the host
than the other and constitutes the major infection.

No. 624] SHORTER ARTICLES AND DISCUSSION 91

Limited geographical areas have been covered in the sur-
veys for cercarie. Two drainage systems of the Atlantic slope,
isolated regions around the Great Lakes, a portion of the upper
Columbia and an isolated region in Wyoming, together with
more widely investigated areas in the Mississippi basin, consti-
tute the localities in which collections have been made. The
entire south, southeast and southwest constitute vast unexplored
areas, the former two of which should yield a great number of
species. In addition, the variation of species of flukes in snails
from one season to another makes it highly probable that many
more species occur in the Mollusca of the areas surveyed than
the records show. Table I shows that one distome species, Cer-
caria megalura, has been found in Goniobasis virginica from the
Atlantic slope, and in Pleurocerca elevatum from the Mississippi
basin; and that C. inhabilis and C. diastropha have been found
on both the eastern and western slopes of the Mississippi drain-
age. On the other hand, none of the species described for
the Bitter Root Valley have been recorded east of the Rocky
Mountains. 5

Records of percentage of infection from larval flukes vary
from a few hundredths of a per cent. for certain cercarie de-
scribed by Sisnitzin in 1911 from the Black Sea to a heavy infec-
tion of every individual of a particular species in a locality.
The lowest infection record for the United States is one per tent.
(C. fusiformis in Physa gyrina). On the other hand, several
heavy infections have been recorded, including three with total
infection. The mollusks most heavily infected are the ubiqui-
tous species, Planorbis trivolvis and Physa gyrina, and the
western species, Lymnaea proxima. In the case of the Planorbis
and the Lymnwa the average heavy infection is caused by dis-
tome cerearie. The heavy infection among the physas is caused
by monostome and holostome larve.

Table II, which summarizes the infection from the host point
of view, shows that Lymnea proxima has the greatest number
of species per habitat. Planorbis trivolvis has been found to be
infected in the greatest number of localities, while Physa gyrina
is the only mollusk to harbor three groups of Digenea. Of the
sixty named species listed in Tables I and II eleven are recorded
from two hosts.

Accompanying the cercariæ in the mollusks are the parthenite
(sporocysts and rediæ) of these cercariz. The cercarie develop
parthenogenetically within these parthenite. Typically, as in

92 THE AMERICAN NATURALIST [Vou. LIII

the life history of Fasciola hepatica, the sporocyst and redia
generations both occur, but in several groups, notably in the
stylet cercariz and the furcocercariæ the redia stage has been
omitted or replaced by another sporocyst stage.

Sporocysts and rediæ have not been sufficiently distinguished.
The sporocyst is an adult which has lost its digestive tube, while
a redia is an adult which possesses both a rhabdoceele gut and a
pharyngeal sphincter. In certain sporocysts the sphincter still
remains, as in C. dendritica. In other sporocysts, as in some fur-
cocercariæ, while no definitely differentiated sphincter is pres-
ent, the anterior end of the sac is muscular, turning in and out
like the finger of a glove. This may easily be mistaken for a
rhabdocele gut.

The cerearie develop within the parthenite and usually at the
time of maturity break out of the parent and work their way
through the tissues of the host into the water. In case no suita-
ble host is at hand in which the larve may continue development
they ordinarily encyst. Groups like the furcocercarie, however,
are not known to encyst. On the other hand, the writer has
found encysted larve of C. biflexa within the larval host and
encysted larve of C. micropharynx even within the parent
sporocysts.

The parthenite of monostome, amphistome and holostome cer-
cariæ are rediæ. Parthenite of certain groups of distome cer-
cariæ are sporocysts and of other groups of distome cercariæ
are rediæ, although some of the records are conflicting. This
shows the need of the accurate determination of the parthenita
of each cerearia, since the parthenita is a distinct generation in
the life history of the species.

In order that the records may not be confusing the writer
proposes the name Cercaria gracilescens for C. gracilis O’Roke
1917, preoccupied by La Valette 1855, and C. minima for C.
minor Faust 1918, preoccupied by Lebour 1912.

In conclusion, the effect of the larva on the mollusk must be
emphasized. It is an observable fact that heavily infected snails
die sooner than uninfected ones. The cause of this mortality is
both the mechanical disruption of the tissues of the infected
mollusk and the pathological changes within the cells of the in-
fected animal. A pathologico-chemical study of this relation-
ship would be of great value to parasitologist and malacologist
alike. ERNEST CARROLL FAUST

UNIVERSITY OF ILLINOIS

No. 624] SHORTER ARTICLES AND DISCUSSION 93

ON RE ATION AND THE RE-FORMATION OF
LUNULES IN MELLITA*

THE ambulacral lunules of the genus Mellita are characteris-
tically developed by the inclusion of reéntrant marginal notches;
except that in M. sexiesperforata these lunules, like the inter-
ambulacral lunule in this and in related genera, is formed by
resorption through the test.2 In a few species the ambulacral
lunules are permanently maintained as open marginal notches,
and Jackson? has called attention to the fact that in Encope
annectans Jackson, the adult interambulacral lunule is repre-
sented by mere dorsal and ventral furrows, not sufficiently deep
to meet and form a hole through the test. In view of the possibly
exceptional character of the formation of the ambulacral lunules
in M. sexiesperforata, note may be made of the manner in which
the lunules of this species are re-formed during regenerative
changes consequent upon natural injury. For in collections of
Mellita comprising half-a-dozen or more individuals of adult
size it is rarely found that every specimen presents a complete
and regular marginal outline. The degree of irregularity is in
most instances not large, but in some cases amounts, at the
deepest point, to a radial deficiency of 13 mm. in a specimen 12
cm. in transverse diameter. I have found no irregularities of
this character in specimens less than 8 em. in transverse diam-
eter. The nature of these deficiencies is such as to suggest that
they have been inflicted by other bottom-feeding animals, pos-
sibly fishes. The wound-surfaces seem, however, to be readily
covered over; and the various degrees of alteration toward a
more perfect outline, exhibited in different specimens, show that
relatively complete restitution is possible.

It is a noteworthy fact that these injuries are almost entirely
confined to that end of the animal which is morphologically the
posterior (cf. Figs. 1, 2, 3). It seems to me possible that this
fact may be understood upon the =a that when burrow-
ing the anatomically anterior end of the ‘‘sea plate,’’ which is
somewhat more sharply pointed than the posterior, is the one

1 a from the Bermuda Biological Station for Research.

2 Agassiz, A., ‘‘ Revision of the Echini,’’ pp. 320-324. Clark, H. L., 1904,
tf Bekinndormns: of the Woods Hole Region,’’ Bull. U. 8. Bur. Fish. for 1902,

pp. 545-576 (p. 565).
3 Jackson, R. T., 1917, Proc. U. 8. Nat. Mus., Vol. 53, pp. 489-501 (p. 494).

FG. 1:

B

Fic. 1. Outline of a Mellita sewiesperforata showing at a and at 8 Lis sta,
in the re-formation of a posterior lunule, At ĝ a furrow, deeper on the cal
surface, marks the region of union of the material from the two niin ria
(x 1.) From life.

Fic. 2. Another case, showing lunules 1 and 5 in process of being closed.
(x 1.) bar living animals.

ee a the goannas of the old lunule-walls has been carried to an

extent eke ae that ¢ the distance between the distal end of the lunule and the mar-
gin of the disc is actually ecenthe than in the case of a lunule normally formed.
Here also the furrow indicated by shading is deeper on the ventral ipara At
B; deformation of a lunule through growth after injury. (x ) From lif

No. 624] SHORTER ARTICLES AND DISCUSSION 95

which is carried ahead.* These animals do burrow beneath the
surface of the sand when the weather is at all stormy, and, if
this assumption be valid, the posterior end might then be exposed
(or even separated somewhat from the bottom) to a greater
extent than at other times. If the process of emergence from the
sand is somewhat different from, or quicker than, that involved
in burrowing, one could understand why the anterior end is
rarely, if ever, damaged, as might otherwise be expected if dif-
ferential exposure of some kind alone determines the incidence
of injuries; there are, of course, other possibilities.

However they originate, the restorative phenomena which these
posterior injuries entail show that the ambulacral lunules of
M. sexiesperforata May, in regeneration, follow a method of for-
mation resembling, in a measure, that adhered to in the normal
development of these lunules by other mellitas. The individuals
herewith depicted in outline (Figs. 1, 2, 3) exhibit several stages
in a process of lunule-completion through the concrescence of
the growing edges of the dise. It is difficult to decide whether
this process is of a specific regulatory character, ‘‘aiming at’’ the
reconstitution of the lunules, or whether it represents merely the
inevitable consequence of ordinary (though accelerated) growth
at the margin of the mellita disc, and is, perhaps, for this reason,
devoid of any recapitulatory significance. An inspection of
Fig. 1 will show that at a there is evident a decided out-bulging
of the disk-margin, at the point of union with the old outline of
the lunule. This out-bulging, seen also at 8 in Fig. 2, and at 8
in Fig. 3, shows definite growth of the tissue toward the opposite
lunule-wall in interambulacrum V. At @ in Fig. 1 an outgrowth
of this type has met and fused with a less extensive outgrowth
from the opposite lunule-boundary ; here, as at a in Fig. 2, it will
be noted that the lateral extension of interambulacral area V is
not confined merely to the margin of the disc, but affects also the
whole lateral wall of the lunule on that side,—provided the injury
be sufficiently extensive—so that closure of the lunule is slow.
If the original disturbance be small, as at 8 in Fig. 1, this and
other similar cases show that reparation may be relatively com-
plete. On the other hand, more extensive injury, as at 8 in Fig.
3, seems to result in ‘‘regeneration’’ which is not so quickly
effective as, for example, in 8, Fig. 1; under these circumstances
the posterior extension of the substance of interambulacrum J,

*Cf. Cole, L. J., 1913, Jour. Exp. Zool., Vol. 14, pp. 1-32.

96 THE AMERICAN NATURALIST [Vou. LIII

not met by growth from interambulacrum V, produces a rela-
tively considerable distortion of the old lunule. Moreover, the
coalescence of the lunule-walls may be carried to a length which
seems greater than it need be, as a a in Fig. 3. For these rea-
sons the idea of a specific regulatory character in the re-forma-
tion of the lunule seems unnecessary.

It would be of interest to observe the effect, upon the course
of regeneration, of an injury deep enough to remove the area
of a lunule down to its proximal border; I have seen no cases
of this kind, nor any in which the interambulacral lunule had
been affected.

Morgan® was unable to find in the literature any mention of
regeneration among echini. The present observations may con-
sequently help to fill a gap at that point, since I am not aware
that any one has previously commented upon the matter, al-
though the conditions described are perhaps well known to other
students of echinoderms. For M. pentapora cases have indeed
been noted in which a lunule, unclosed, extended to the periphery
of the disc; but these instances have been referred either to
accidental injury or to delayed growth, and ‘‘repair’’ phenomena
seem not to have been observed. In M. sexiesperforata regenera-
tion (of a sort) does undoubtedly occur, giving evidence of a
respectable degree of plasticity in a stony structure where such
might not be expected.

W. J. CROZIER

DYER ISLAND,
BERMUDA

5‘‘ Regeneration’? (1901), p. 105.

THE
AMERICAN NATURALIST

Vou. LILI. March-April, 1919 No. 625

AN EARLY PAPER ON MAIZE CROSSES

PROFESSOR HERBERT F. ROBERTS

KANSAS STATE AGRICULTURAL COLLEGE

In the writer’s opinion the paper of MeCluer, entitled
‘‘Corn Crossing,” has lacked adequate appreciation by
later investigators. This paper is referred to by East
(3), Shull (7) and Collins (1) in various articles, but only
with reference to McCluer’s observations on the superior-
ity of F, hybrids in point of yield, as compared with their
parents. The other matter of genetic interest in the
paper seems to have attracted little attention, perhaps
because of the more extensive earlier experiments (’89-
91) of Kellerman and Swingle (4), and the more impor-
tant later ones of Correns (2).

However, with respect to the superior yield of F, hy-
brids, McCluer’s paper exceeds in interest the more fre-
quently quoted ones of Morrow and Gardner? in respect
to the wider variety of types selected for crossing, and
especially in the fact of the progeny of the F, plants hav-
ing been followed out in the yields of the F, generation,
which, in all but a very few cases, were found to be dis-
tinctly less than the yields of the F, plants. This seems
to have been, historically speaking, the first demonstra-
tion of the inferiority in yield of F, segregates, as com-
pared with first generation hybrids. MeCluer’s experi-
ments involved a wider range of types of maize than did
those of Morrow and Gardner, and comprised dent, sweet,

1 Til. Ex. Sta. Bull. 21, May, 1892.

2 Ill. Ex, Sta. Bull. 25, 179-80, and 31, 359-60.

97

98 THE AMERICAN NATURALIST [Vou. LIII

pop and soft corn varieties in eighteen different crosses.
He was thus enabled to make observations on the inherit-
ance of characters other than those resulting in yield.

In 1889 MecCluer began his hybridization: work with
corn, crossing a number of strains of dent maize, without
at first, however, making crosses between varieties of
different colors. In addition he utilized the following
characters in thirteen crosses made in 1889, and in five
crosses made in 1890 obtained in the former year 36 and
in the latter year 158 ears.

Sugary endosperm starchy endosperm.
yellow endosperm and the reciprocal.
corneous endosperm and the reciprocal.
sugary endosperm.
purple aleurone

mM
S
3
Sy
4
o
=
a
©
=
S
B
KAX EX

Non-colored aleurone

Expressing these crosses in the following notation,
according to presence and absence, for convenience, with-
out regard to the actual gametic composition of the pa-
rents, which of course can only be inferred, we have:

Yellow endosperm.

Colored aleurone
Non-colored aleurone.
Corneous endosperm.

as ars &
n
aga
D
a
oO
z,
3
g

McCluer’s crosses then classify according to the for-
mulas on the following page.

Leaving it understood that this is simply a classi-
fication in shorthand formulas, of the visible characters,
without predicating their gametic composition, which was
unknown to McCluer, since he far antedated the days of
pure lines, Mendelism and factorial analyses, we are
struck by the number of factors with which he experi-
mented, and with his clean observations on the results.
Being a horticulturist, he was led to be interested in these
various types of maize, to which Morrow and Gardner,
as agronomists, gave no attention.

No. 625] AN EARLY PAPER ON MAIZE CROSSES 99

Cross Ps Lads Variety ap ph a Variety
ters
1 abe Mammoth X ~ ABe Leaming
2 abc Triumph x Abe
3 abe row X ABe
4 abe Mammoth X Apo Golden Coin (Sweet)
5 abe Triump se Abe
6 abe 8-row x Abo oe us
7 abe Mammoth o ADe Stowell’s kviki
8 abc Triumph X Abe
9 abe 8-rowed X.. abe ee
10 | abe White Dent (unnamed) X ADe Queen’s dü en Pop
11 | ADe Queen’ 8 Golden Pop Ae White De = Capeepied:
12 ADe po nE Bla ck Mexi
13 | ABc White Dent (unnamed) X abC
'14 | ABc Brazilian Flo x Abc Gold Coin (Sweet)
15 | ADe | Pearl (pop) X ADS Queen’s Golden (Pop)
16 aBe Burr’s White eye XO ae Brazilian Flour
17 aBe Yellow Dent variety Xo KBC White Dent variety
x

18 aBe White Dent erin be -aBa Yellow Dent variety

McCluer found no xenia effect to be produced, where,
as he says, ears ‘‘of the same color” but of different types
are crossed.

The typical ear of Stowell’s Evergreen differs very decidedly from
typical ears of either 8-rowed, Triumph or Mammoth, but the ears
produced by pollen of Stowell’s on either of the others, did not differ
from the female type in any way, more than did many ears left to be
fertilized naturally.

In other words, he observed that the ‘‘maternal’’ tis-
sues beyond the endosperm were not affected by the
crossing.

McCluer TR that in F, ears, of crosses between yel-
low and non-yellow endosperm, the dominant yellow of
the F, kernels was never as dark as in the yellow parent,
whether the latter were the pollen or the seed parent.
This fact was observed in crosses 1, 2, 3, 10, 11, 13, 17
and 18, Since McCluer remarks that the effect was not
uniform in yellow dent X white dent crosses and their
reciprocals, he may in these cases have unconsciously run
across the phenomenon of two yellows, as reported in

911 by East and Hays (3), pp. 46-56. It is probable,
however, that most of the instances were cases of the
heterozygous yellows, being lighter than the homozygous

100 THE AMERICAN NATURALIST [Von LHI

yellows, as reported by East and Hayes, loc. cit., pp. 55—6.
In McCluer’s crosses, 14, 17 and 18 at least, he was evi-
dently working with endosperm color factors, the be-
havior of which was identical with those reported by East
and Hayes in their crosses as given in the citation above.

McCluer made a considerable number of observations
on xenia, but remarks (italics mine), ‘‘The results ob-
tained from planting crossed seed have been of more im-
portance than the immediate efect of crossing, not so
much in themselves perhaps as in the conclusions which
may be drawn from them.” Such a point of view could
only have been arrived at by one with something of an
instinct for genetic studies.

McCluer remarks upon the great uniformity of what
we should call F, hybrids. He says:

Of 142 plots planted with sweet corn, pop-corn and their crosses, it
is safe to say, there was as much uniformity in any one of the crossed
plots, as in any, and very much more than was found in most of the
plots planted with pure varieties.

Some interesting notes were made as to the character-
istics of some of the F, plots. For example, it seems
that the plots in which Leaming was used as the pollen
parent decidedly resembled that parent; that in crosses
between Queen’s Golden X White Dent, the F, plants
resembled the pollen parent, whereas in the reciprocal
cross, the plants were intermediate between the two pa-
rents. An interesting result came from the cross between
Queen’s Golden and Pearl pop-corn. The stalks were
intermediate between the parents, but larger than the
average of the two parents. This characteristic extended
to:the growth of the cob, so that the F, ears were dis-
tinctly larger than the average of the ears of either of
the parents—a fact very well illustrated in Plate 2 of the
bulletin. Plots of F,, hybrids between White Dent X
Black Mexican, decidedly resembled the white dent. An
extraordinary result seems to have been obtained in this
cross. The F, seeds—i, e., the seeds of the white dent
ear pollinated with Black Mexican pollen, show the usual

No. 625] AN EARLY PAPER ON MAIZE CROSSES 101

dominance of purple aleurone in the F, kernels. But the
starchy character, ordinarily completely dominant in F,
seeds of starchy sugary endosperm crosses, is not dom-
inant in all the kernels. So far as the F, ear illustrated
on Plate 1 indicates, on which there are four tolerably
complete rows in sight, there is a ratio of wrinkled to
smooth kernels of 73:50, or approximately 1:1, which
would go to indicate that the seed parent was probably
heterozygous as to starchy endosperm. The wrinkled
seeds from this ear produced ears, to judge again from
the plate, for there is no detailed description, that were
pure wrinkled in their kernels, whereas the smooth ker-
nels from the F, ear produced ears on which both smooth
and wrinkled F, kernels were borne. Three such ears are
illustrated. On each of these ears three complete rows of
kernels are visible in the illustration. Counts of these
kernels, as nearly as they can be made from the illustra-
tions, show: In ear No. 3 (i. e., the ear showing the im-
mediate effects of the cross and bearing the F, kernels)
the ratio of smooth seeds to wrinkled seeds is as 73:50.

It was eight years later that the papers of Mendel were
rediscovered, and at this time no scientific knowledge of
the genetic behavior of corn existed at all. MeCluer ob-
tained, however, very definite evidence that the ears
“borne by hybrid corn plants grown the first year from
the cross,” as he puts it, or as we should say to-day, plants
of the F, generation, were larger on the average than the
average ear borne by the parents, and that the yield was
greater. Taking McCluer’s tabulations of his results on
p. 97 of Bull. 21, and revising its notation to correspond
with present usage, we have the data given on page 102.

From this early experiment the result of crossing, so
far as the yield of the F, generation is concerned, is fairly
well indicated, since in fourteen cases out of eighteen the
F, hybrids yield more than the average of the two pa-
rents, although in only seven cases did the yield of the
F, hybrid exceed that of both the parents. McCluer also
emphasizes the inferior condition of the self-fertilized
plots.

102 THE AMERICAN NATURALIST [Vou. LIII

Av. Wt. | Wt- (Oz.) of
Wt. (Oz.) of | Wt. (Oz.) of (Oz.) 10 10 Ears
Cross 10 Ears of | 10 Ears of Ears of Borne by
the Male | the Female! the Two a = ite

Parent Parent Parents (ep ieee

Queen’s Golden (1) X White Dent . 34.50 81.00 57.75 76.00
Whi X Queen’s Golden....... 81.00 34.50 157.75 64.00
Queen’s Golden X Black Mexican (2).| 34.50 36.00 35.25 47.50
mmon Pearl (1) X Queen’s Golden.| 27.50 34.50 ‘31.00 42.00
Ma J X Geaming (3) iana 61.50 87.50 74.50 91.00
Mammoth X Lea S AORE R Sirie ee St E 61.50 87.50 74.50 82.00
Amme h x Deonming sso eaea 61.50 87.50 74.50 80.50
Triumph X Leamin p R pwa oc orae a ee 46.50 87.50 67.00 83.00
iabe tom ód (2) X-Lëaming iris naa 41.00 87.50 64.25 72.00
Brazilian Flour GN x ‘Gold Coin (2).| 39.00 63.00 57.00 78.00
White Dent * Black M rie MAERA A 81.00 36.00 58.50 51.00
Eight-rowed X Stowell’s Seaece (2)| 41.00 57.50. 49.25 47.00

Triumph (2) X posed: vergree 46.50 57.50 52.00

Mammot ell’s Evergree 61.50 57.50 59.50 61.00
Gold Coin X Stowell’s Evergreen. . ee 62.50 57.50 60 00 62.50
Trivmph X Gold Coin) ive. RS 46.50 62.50 54.50 58.50
EHight-rowed X Gold Coin........... 41.00 62.50 51.75 56.00
Eight-rowed X Gold Coin........ Su She 62.50 51.75 58.00
eric a eas 50.50 | 63.30 | 57.20 | 64.50

(1) Pop corn, (2) sweet corn, (3) dent corn.

Plots grown from self-fertilized seed, were in most cases notably in-
ferior in size and vigor to the plots grown from crossed seed, or from
seed simply selected. The table does not give so convincing an illus-
tration of the bad effects of self-fertilization, as the plots themselves
did when growing, or as the corn did when husked and thrown into
piles. One plot from self-fertilized seed had nearly half the stalks
deformed in such manner that instead of standing up straight, they
turned off nearly at a right angle, at or near the joint where the ear

plot from self-fertilized seed, nearly all the tassels were abortive. All
the plots from self-fertilized seed produced a greater proportion of
barren stalks, and of poorly filled ears, than the plots of the same varie-
ties, either from crossed seed or from seed naturally fertilized. The
table giving the weight of ten selected ears of corn from self-fertilized
seed, and of ten ears from crossed or seletted seed, does not give a cor-
rect idea of the inferiority of the corn from the self-fertilized seed,
because it does not take into account, either the greater proportion of
barren stalks, or of small poorly filled ears (pp. 96 and 98).

The results of this experiment at the time simply led
to the conclusion that continued selection of corn, lead-
ing to a certain amount of inbreeding, was likely, like close

No. 625] AN EARLY PAPER ON MAIZE CROSSES 103

fertilization consciously practised, to lead to ‘‘deteriora-
tion,’’ and that cross fertilization, as it occurs ordinarily
in corn, is desirable for the best results. No suggestion
is offered by McCluer as to utilizing this fact in a practi-
cal way. It remained for Messrs. Morrow and Gardner,
also of the Illinois Station, to derive this conelusion from
their experiments. In Bulletin 25 of the Illinois Station,
pp. 179-180, results are given of crosses made between
dent corn varieties exclusively, which, while less exten-
sive and varied than MecCluer’s, are confirmatory of his
experiments. The following table, adapted from Bull.
25, p. 180 (1893), presents these results:

Variety Bu. Air-dry Corn per Acre
Barre Waite OS ees tae ee eae 64.2
OranDorry cics ne 65 S ad 2G a oe 61.6

Average viia. vet. SiS esee és es 62.9

CV ORS So See ae a et eee 64.1
Burr's Witte s- E E R wane 64.2
HORNS REAGROUOR eeoa a en aAa haau 79.2

r adis u Or rer e TE ea ee ak TLI

CORA E O PAA es is E E E 73.1
KONE se es elena ibaa 73.6
Goldon: Beaty il ee ne or 65.1

AVOPATO c.s A Wee et webs 69.3

eyes sis eee ie We Vi vies 86.2
Champion White Pearl 54.05.0505 «vices 60.6
AF Or aera aoa bee rere piel. Burg pair gers 73.6
Averea ee oe hh cies 67.1

PORE 555k vac oa ee ed ate 76.2
Bure’ sa. We 6 oi tc a oe enemies 64.2
Be -zi ee rr et ere 58.4
ANOTADO FSi. Ae aN ALESE oE 61.3

Crone Foes sak cho oa ieaie T 78.5

In each of the above cases the yield from the cross ex-
ceeded the average yield of the two parents, although not
in all cases exceeding that of each parent.

In Bulletin 31, pp. 359-60 (1894), the result of Morrow
and Gardner’s second experiment in crossing corn is
given.

104 _ THE AMERICAN NATURALIST [Von. LIII

corres Wito Pearl. oon Sy oie
Bo WANE oc Sates eset Tes Pee E 38.6
ERER Bate ary, E A E oe 38.0
RR ss REE EEE E ARE E HR a 28.4
N PE RE EE ll E oa EE E 34.6
rag s ik ie Sed Oke AS N E T 38.6
VOTRE aE oor een et es ere 36.6
PORES.) Poke Ge Cone oe iea we rae | oh 41.7
Edmund a clk a. 20s sce sa Pee hs ees 28.3
Murg osor en ince ec See eee 35.7
AVOVARR plone ee et ee pas oo 32.0
PORE Ge Tk eh tg Cer ue as ue 41.4
AAW E ng Sek es PES OSES, e Bie 28.3
Burr 8 WIRE eh ak co hes oie a 38.6
POU snk oe os E oR 33.5
37.8

ee RAAR A

In three out of the four cases above the cross out-
yielded the average yield of the two parents.

Some observations were made by MecCluer, (p. 86), as
to effect of crossing on the number of rows of kernels,
the results being an approximately intermediate condi-
tion in the F, hybrids with respect to this character.

arents No. of Rows

Be te e ere re ere ee ere ere ee 18-24
Mammoth Sweet) 255. hs eee ea wae 12-16
si MUS es See ee et ee S 14-18
me pupae Weer ne ee ee s 18-24

ight Towo SWOT cis ow a aos Pee eee 3
FE BY DRIOG es re ites tee ee ais 10-14
OUT 8 0 ay eas eee oe eee ae eee 18-24

T aa a aa a leds es eos wee ee
I DY DPI e a es ce E tee eee ee 10-16

MecCluer remarks upon the difference in reciprocal
pop-corn—dent-corn crosses, to the effect that when the
pop corn was used as a pollen parent, the F, kernels were
more flinty than when the dent corn was used as the pol-
len parent.

So far as the writer knows, McCluer is the first person
known to have made a cross between two different types
of corn, who paid close enough attention to the results of
such a cross to lead him to illustrate the parent ears, the

105

S

A

SE

CROS

4
vi

R ON MAIZE

7
vi

ARLY PAPE

7
a

AN E

No. 625]

tat t
HUN a
tod j

aojguid iinssicscautiiaibonssaunuss ive
Rat S Satan y
SHEE HEULI 119°
EATON TTTTLTETT 8D 8}
RETTE WHT un TINS

reat ak EEES

TERREN
Moreh Dr CATT
pe ness tee
+ ‘te ore sont Ss FAS
? F Shc Giese sO deans oe TT TTT
oe erases SES eee ae) ee

ee inde P

yey Yeeeah
serei

pre ere i

as

aes
ies oes

` ae as YTY aa i < ga
ATT cos T ELAT T ‘
eee? (esc fe ake yes Pee Pens J

peed oy heehee oT ae aay

Sebbratee ttt the peed tar os
CER PAPOEA TTT EEE ETTE
ases tres cried TAT AAAA
Urkers ra PE PAASA

ae

2.

Fig.

106 THE AMERICAN NATURALIST [Vou. LIII

ears produced as the result of the cross (F,) and the
second generation hybrid ears (F,), together for compari-
son (Figs. 1 and 2; MeCluer’s Plates 1 and 2). Each of
these is a dent-sweet-corn cross, and the results, both of
F, dominance and of segregation in the F, ears, is very
plainly shown. ,In MecCluer’s Plate 4 the results of
segregation are shown, so far as the reappearance of
parental types is concerned. Of course, in all these
crosses, it must be remembered that the parental types
were not selfed strains, but were undoubtedly heterozy-
gous for some of the facture under observation. This is
shown in the white corn ear, coming out of a cross be-
tween Leaming (yellow dent) and Mammoth Sweet.

The ears shown as types of the varieties used in crossing are se-
lected typical specimens of the varieties, and the ears shown as grown
from the crossed seed are typical of the cross-bred corn (p. 95).

McCluer makes the penetrating remark regarding the
production of F, seeds that

The self-fertilized ears showed the same modification of kernels as
those naturally fertilized, proving that each kernel of the crossed corn,
had in itself the power to produce both sweet and dent corn (p. 95).

In the writer’s opinion, this is the most remarkable ex-
pression wpon the nature of heterozygosis made before
Mendel’s time.

The reappearance of parental types is 3 reterred to as
follows:

Where the parent varieties were widely different, as in the crosses
between sweet and dent, the progeny has tended strongly to run back
to the parent forms, while at the same time taking on other forms dif-
ferent from either (p. 95).

A further indirect comment on the superiority in size
on the part of F, hybrids is seen in McCluer’s statement
that

Nearly all the corn grown a second year from the crosses is smaller
than that grown the first year, though most of it is yet larger than the
average size of the parent varieties (p. 96).

McCluer comments emphatically on the inferior con-
dition of the self-fertilized plants and remarks:

<

No. 625] AN EARLY PAPER ON MAIZE CROSSES 107

he table giving the weight of ten selected ears of corn from self-
fertilized seed, and of ten ears from crossed, or from selected seed, does
not give a correct idea of the inferiority of the corn from self-fertilized
seed, because it does not take into account, either the greater propor-
tion of abortive stalks, or of small and poorly filled ears (p. 98).

The fact is noticed that some varieties, when crossed,
give rise to plants of increased size, while others do not.

Among other incidental matters, MeCluer calls atten-
tion to the necessity for ‘‘A more perfect knowledge of
the development of the races and varieties of corn,” and
wisely remarks regarding the farmer’s part in corn
breeding:

In the production of new varieties by crossing, it will seldom be
desirable to cross two varieties that are very widely different from each
other. Itis probable that, on the whole, selection, with occasional par-
tial changes of seed, will give more permanent as well as more satis-
factory results for the general farmer, than would the continual cross-
ing and breaking-up of well fixed types; though there does seem rea-
son to believe that the crossing of such distinet and well-fixed types,
will, for the time being at least, give larger corn and better yields
(p. 98

From MecCluer’s observations on the results in the sec-
ond generation of the hybrids he comes to the following
intelligent conclusion:

This work gives us a clew to the relative prospects of improvement
in other lines by ecross-breeding. A variety or type that is strongly
fixed is more apt to transmit characters than one poorly or not at al
fixed. If we should try to improve corn by erossing the product of
two of these cross-bred groups of corn, we should expect to get as a
result a few superior ears, with a very large proportion of inferior
ones. Even in our well-selected varieties that have been picked for
years with reference to given points of excellence, the tendency to run
back to inferior forms is so strong, that the grower would save hardly

one-tenth of his crop for his own seed. If our well-selected varieties
deteriorate thus, when constantly and carefully selected, two varieties
that have been long selected for opposite or widely different qualities,
must, when crossed, tend to neutralize most strongly the very traits
which we have, with so much pains, brought out and maintained.

If, on the other hand, the varieties crossed have long been selected
on the same or very similar lines, there seems to be no reason wh
casional crossing will not tend to fix more strongly the desired char-
acters.

108 THE AMERICAN NATURALIST [ Vou. LII

Here, of course, MeCluer quite naturally overlooks the
fact of dominance, and adheres, although with a more
rational reason than most plant breeders of his time, to
the idea of fixation of type through the effects of selec-
tion. McCluer, however, here as throughout his paper,
shows the inherent instincts of a geneticist, and his paper,
although an obscure contribution to the literature of plant
breeding, deserves special notice on that account.

LITERATURE CITED
1. Collins, G. N.
1910. The Value of First Generation Hybrids in Corn. Bur. Pl. In-
dustry, U, S. Dept. Agric. Bull. 191, Oct.
. Correns, C. :
1899. Enana über die Xenien bei a Mays. Berichte der
schen botanischen Gesellschaft, 17: 410.
3. act iia i and Hayes, H. K
911 TEUA in Maize. Conn. Ex. Sta. Bagy 167, ope

bo

4, Tiea ak William A., and Swingle, Walter
1888. Experiments in Cross-fertilization k ER First Kansas Ex.
Sta. Rept., pp.
1888. Experiments in Crossing Naristiks of Corn. Second Kansas Ex.

a. pp. 288-355.
5. tage oy George W.
Corn Crossing. Ills. Ex. Sta. 25 21, May.
sn si E., and Gardner, Fran
sh Experiments with Corn. "oy Ex. Sta. Bull. 25, April.

=

~J

. eee See
he ees of Maize. AMER. NAT., 44: 234, April.

HYBRIDS AMONG SPECIES OF SPIROGYRA!

PROFESSOR EDGAR NELSON TRANSEAU

Oxnto STATE UNIVERSITY

Consucation between filaments of different species of
Spirogyra have been reported by several students of the
alge. Bessey (1884) reported and figured zygospores
formed by the crossing of S. majuscula and S. protecta.
He noted that the spores formed by the cross in either
direction corresponded to the spore type of the female
filament. This has been confirmed by all subsequent ob-
servers and we shall see later that this is a necessary
result of the process of fertilization as it occurs in Spiro-
gyra. This cross is of particular interest because the
vegetative cells of majuscula have plain end walls, while
those of protecta have replicate end walls. Consequently
there is nothing in the physiology of these two species,
representative of the two divisions of the genus, that in-
terferes with conjugation. Wolle (1888) figures a cross
between S. maxima and 8. ? nitida under the name of
S. maxima var. inequalis. West and West (1898) figured
a cross between two of the smaller species of Spirogyra
but did not give their names. Andrews (1911) figured
and described a cross between S. crassa, one of our larg-
est species, with several chromatophores in each cell, and
S. communis, one of the smallest of our species, with a
single chromatophore. Here again the spore formed re-
sembles that normal to the female filament. Evidently
differences in size, in number of chromatophores, shape
of zygospore, and character of the end walls of the vege-
tative cells are not impediments to crossing. In the col-
lections from central Illinois which I have studied I have
found hybrid zygospores formed between three pairs of

1Papers from the Department of Botany, The Ohio State University,
o. 104. :

109

110 THE AMERICAN NATURALIST [Vou. LII

species: S. communis X S. varians from one locality, S.
varians X S. porticalis from two localities, and S. mas-
ima X S. submaxima from one locality.

The phenomenon of hybridization in this genus is evi-
dently quite rare, as shown both by the small number of
references to it in the literature and by the few cases
that have come under my observation. From the collec-
tions made in Illinois, Massachusetts, Michigan, Ohio and
New York, I have 854 records of conjugating Spirogyras
and only five records of conjugation between different
species from four localities, all in central Illinois.

In this paper I wish to record, not only the finding of
conjugation between species, but what seem to be the
progeny of such crosses. It has been found impossible
to cultivate these forms in the laboratory so that there
is no experimental proof of their origin. Nevertheless,
they have been found associated with filaments that were
crossing and with filaments of the two parent species that
were conjugating normally. There were few other spe-
cies present in these collections, and there are no species
in central Illinois that could possibly be confused with
them. Under the circumstances it seems impossible to
account for the strange mixture of forms in these five col-
lections except on the basis of a hybrid origin.

Spirogyra varians X S. communis

In collections taken from Campus Creek, two miles
southwest of Charleston, Ill., during the latter part of
May, 1913, specimens of typical S. varians and S. com-
munis occur together with filaments that resembled one
or the other of these species but whose dimensions ex-
clude them from these species. In going over these col-
lections in 1915 filaments of varians were found conjuga-
ting with communis, and hybrid zygospores were found in
both kinds of filaments. This suggested that the unnam-
able forms were the progeny of hybrid zygospores.

S. varians has vegetative cells .30-40 » X 30-120 p, a
single chromatophore making one to five turns in the cell.

No. 625] HYBRIDS AMONG SPECIES OF SPIROGYRA 111

The sporiferous cell is usually inflated on the conjugating
side and the spores are frequently placed obliquely. In
conjugating filaments cells that fail to mate usually be-
come greatly distended. The spores are ellipsoid, 32-40 p
X 90-100 (Fig. 1). This species is highly variable, but

Fic. 1. Typical 8. varians showing spore ferm, sporiferous cells inflated on
the conjugating side, and an inflated sterile cell. Camera lucida drawing, same
seale as the succeeding figures.

I have examined many collections containing it and have
seen none that resemble the hybrids.

S. communis has vegetative cells 18-26» 35-90p, a
single chromatophore making two to four turns in the
cell. The sporiferous cell is cylindrical and the spores
are placed longitudinally in the cells. Cells in conjugat-
ing filaments that fail to mate usually remain cylindrical.
The spores are ellipsoid, 19-26 » X 35-90» (Fig. 2).

ee

Fic, 2 Typical S. communis ganhe ge of zygospore, sporiferous cell and

The collections in question showed in addition to the
typical filaments others with the characters of varians,

112 THE AMERICAN NATURALIST [Vou LIT

-

Fic. 3. 8. varians x 8S. communis Petik ro A of communis, otherwise
resembling varian

but with dimensions similar or near to communis (Fig. 3)
and still others with the form characters of communis,

ut the dimensions of varians (Fig. 4). The figures are
camera drawings all made from Coll. No. 1877, May 27,

Fic. 4. 8. variansxS. communis showing dimensions of varians, otherwise
resembling communis.

Spirogyra varians X S. porticalis

In collections from a small stream that flows under the
Clover Leaf R. R., just east of the station at Lerna, Ill.,
collections made dariak May, 1913, and April, 1914, Spiro-
gyra varians was found conjugating with S. porticalis.
Accompanying the typical varians and porticalis fila-
ments were fruiting filaments which could not be placed
satisfactorily in either species, but which possessed vari-
ous combinations of the characteristics of both species.

No. 625] HYBRIDS AMONG SPECIES OF SPIROGYRA 1138

In collections from Cossel Creek, about one mile west
of Charleston, Ill., made during May, 1914, and April,
1915, a similar mixture of forms was found associated
with filaments of typical varians and porticalis. It was
the discovery of these latter collections in 1915 that led
me to go over the collections from Campus Creek, Lerna,
and the previous collections from Cossel Creek. In all
eases the forms seemed quite explicable on the assump-
tion that hybridization had occurred in previous years as
it was occurring when the collections were made and that
the progeny of the hybrid zygospores showed various
combinations of the characteristics of the parent species.

FERTILIZATION IN SPIROGYRA

Fertilization in Spirogyra takes place by the fusion of
two gametes through a tube formed by the union of
emergences from two adjoining cells in the same or dif-
ferent filaments. The zygospore matures its wall within
three or four days after the passage of the gametes. -

In the large species, S. ellipsospora, I have watched
the movement and fusion of the gametes. The male
gamete withdraws slightly from the gametangium wall;
its chromatophores disintegrate and the whole gamete
appears to have a foam structure. Opposite the tube a
small lobe develops and this moves through the tube to
the female gamete. On reaching the surface of the fe-
male gamete the cytoplasm of the female gamete at the
point of contact spreads, drawing its chromatophores
apart. The male gamete penetrates the cytoplasm and at
the end of the process lies entirely inside the female
gamete. Its chlorophyll then turns brown and stains the
eytoplasm, so that it can be seen through the maturing
spore wall for several days. During this period it gradu-
ally spreads out and coalesces with the cytoplasm of the
female gamete.

Since the cytoplasm that secretes the zygospore wall
is thus only the cytoplasm of the female gamete, its form
and markings are determined entirely by that gamete.

114 THE AMERICAN NATURALIST [Von. LII

All the recorded facts concerning hybrid zygospores are
in harmony with this observation.

The male and female nuclei, however, do not fuse until

some time later, perhaps two to four weeks. The charac-
ter of the zygospore is therefore entirely dependent upon
the female parent. This is well brought out by all the
published figures of hybrid zygospores as well as by my
own observations. After the formation of the fusion
nucleus various observers—Chmielewski (1890), Trondle
(1907), Karsten (1909) and Kurssanow (1911) (Zyg-
nema)—report the occurrence of a double mitosis of
which the second division is heterotypic. This results in
the formation of four nuclei, of which three degenerate
(Trondle, 1911). The degeneration of three of the nuclei
has been observed also by Kurssanov in Zygnema. The
remaining nucleus becomes the final nucleus of the zygo-
spore.

In a hybrid zygospore, therefore, the first fusion nu-
cleus would contain the hereditary factors for the alter-
nate characters of both species; in the subsequent reduc-
tion division and degeneration of three of the nuclei the
final zygospore nucleus would contain one of the several
possible combinations of these hereditary qualities. The
vegetative filament derived from a hybrid zygospore
would present some combination of its vegetative charac-
ters, such as cell dimensions, number of chromatophores
and character of the end walls. The factors for spore
characters would not become visible until it conjugated,
and then only the characters carried by the female fila-
ments would become visible since the spore characters are
entirely matriclinal. In the adjoining diagram (Fig. 5)
the zygospores are figured in typical forms. The two
characters most important in separating the species are:
(1) The dimensions of the filaments (varians averages
about 33, communis about 22); (2) the lateral inflation of
the sporiferous cell in varians in contrast with the cylin-
drical sporiferous cell of communis. If we represent the
factor for the dimensions of varians by A and of com-

>

No. 625] HYBRIDS AMONG SPECIES OF SPIROGYRA 115

munis by a, and the factor for the inflated sporiferous
cell of varians by B and cylindrical sporiferous cell of
communis by b, then the fusion nucleus in the zygospore
would have the composition 4aBb. At the subsequent
reduction division these four characters might be distrib-
uted in four different ways: AB typical varians; Ab vari-
ans dimensions and communis type of sporiferous cell;
aB—communis dimensions and varians type of sporifer-

AaBb.

SPAS
eleri

Fig. 5. Diagram illustrating pi foiatta progeny of a cross between 8. varians
d 8. communis.

ous cell; and ab—typical communis. All four of shoe
possibilities are represented by filaments in Collection
1877 from Campus Creek. Furthermore, the fertile cells
of individual filaments and the zygospores formed are
similar throughout a particular filament, as would be ex-
pected on the basis of the origin of the final zygospore
nucleus. This hypothesis, therefore, accounts for all the
facts at present known concerning this series of collec-
tions.

The forms occurring in the two series of collections

116 THE AMERICAN NATURALIST [Vou LIIT

from Lerna and Cossel Creek can be similarly accounted
for. In this case, however, the two species, S. varians
and ©. porticalis, evidently differ in three characters.
The characteristics of varians have been given above.
S. porticalis (Fig. 6) has vegetative cells 40-50» x
66-200 containing a single chromatophore, making
three to four turns. The fertile cells are cylindrical and
the zygospore is ovoid, 38-50» X 50-83». It, therefore,
differs from S. varians in dimensions of the vegetative
cells, in the absence of inflated fertile cells, and in the

Fic. 6. Typical 8. porticalis showing single chromatophore, ovoid spores and
cylindrical sporiferous cells.

ovoid form of the zygospore. In both series of collections
all of the eight possible combinations of these three char-
acters occur.

In the accompanying diagram. (Fig. 7) the dimension
factor for varians and porticalis are represented by A
and a, respectively. The factor for the varians type of
fertile cell by B, and for the porticalis type by b. The
factor for the ellipsoid spore by C and for the ovoid spore

y c. The nucleus formed by the fusion of the two
gamete nuclei therefore contains all the factors (Aa, Bb,
Cc). The final zygospore nucleus contains any one of
eight possible combinations: ABC, typical varians; AbC,
varians dimensions and spore form, with porticalis fertile-
cell form; Abc, varians dimensions, with portiealis fertile-
cell form and spore form; ABc, varians dimensions and

No. 625] HYBRIDS AMONG SPECIES OF SPIROGYRA 11T

AbG aBe

ABc aBC.

doo

a em

Fic. 7. Diagram to illustrate hybrid progeny of a cross between S. varians
and 8. porticalis.

118 THE AMERICAN NATURALIST [Vou LIII

fertile-cell form, with porticalis spore form; abc, typical
porticalis; aBc, porticalis dimensions and spore form,
with varians fertile-cell form; abC, porticalis dimensions
and fertile-cell form, with varians spore form; aBC, por-
ticalis dimensions, with varians fertile-cell and spore

rm.

The female filaments derived from these several zygo-
spores should show these eight possible combinations
when they fruit. These forms have all been found in the
two series of collections in each of two successive years,
and there seems to be no question but what they are the
products of hybridization and segregation. Since the in-
heritance is entirely matriclinal, segregation occurs in
the first generation.

As to the relative numbers of the several types it is
impossible to count filaments in a collection. Theoreti-
eally they should be present in about equal numbers if all
the filaments are of hybrid origin. In all the collections,
however, there were filaments conjugating in the usual
way, so that any attempt at counting filaments would be
useless even though it were practically possible.

SuMMARY

1. Hybridization between Spirogyra communis and S.
varians and between S. varians and S. porticalis have
been observed. The forms probably derived from these
crosses have also been found.

2. Hybrids are exceedingly rare among species of
Spirogyra.

3. They have been observed only in a few species.

4, Hybrid zygospores may be formed between species
even though they have very different vegetative and spore
characters.

5. The nuclei derived from the two gametes do not fuse
until after the zygospore wall matures. The form of the
zygospore is determined entirely by the female gamete.

6. The fusion nucleus of a hybrid zygospore contains
factors for all the various characters of both species.

No. 625] HYBRIDS AMONG SPECIES OF SPIROGYRA 119

When the reduction division takes place these factors are
segregated in various combinations in the final spore
nucleus.

7. When the filaments derived from hybrid zygospores
fruit their hybrid character becomes evident in the dimen-
sions of the filament, the character of the sporiferous cell
and the form of the zygospores.

8. Inheritance is matriclinal and therefore the segrega-
tion becomes evident in the first generation.

LITERATURE CITED

Bessey, C. E
1884. Hybridism å in Spirogyra. AMER. NAT., 18: 67.
Wolle, Francis.
1887. Freshwater Alge of the United States. P. 218. Plate

CXXXVIII, Figs. 5 and 6, and CXLII, Figs. 5 and 6.
West and West.
1898. Ses hte age on the Conjugate. Annals of Botany, 12: 43;
e V, Figs. 70 and 71.
Andrews, F. =
1911. Conjugation between Two Different Species of Spirogyra. Bull.
Torrey Botanical Club, 38: 299.

Overton, E.
1888. Über bege eee e bei Spirogyra. Ber. d. Deutsch.
Botan. Ges., 6: 68-72.
Klehbahn, H.
1888, Uber die Zygosporen einiger Conjugaten. Ber. deutsch. Botan.
Ges., 6: 160-166.

Chmielewski, V. F.
1890. Materialen zur Morphologie und Physiologie des Sexualprozesses
bei niederen Pflanzen. Arbeit. Ges. der Naturf. d. Charkower
Univ., 25
Troéndle, A.
1907. “poe die — und Keimung von Spirogyra. Botan.
Zeitung, 65:
1911. Uber ye Hoduktonsteilung in den Zygoten von Spirogyra und
über die Bedeutung der Synapsis. Zeitschrift fiir Botanik,

3: 593-619.
Karsten, G.
1908. Die a ae der tek von aes jugalis Ktzg.
Flora, 99:
Kurssanow

1911, ibe Befruchtung, Reifung, und Keimung bei Zygnema. Flora,
104: 65-84.

SYNTHETIC PINK-EYED SELF WHITE
GUINEA-PIGS!

DR. HEMAN L. IBSEN

INTRODUCTION

In the fall of 1914 Professor W. E. Castle kindly sent
to the Department of Experimental Breeding, Univer-
sity of Wisconsin, five guinea-pigs representing some of
the rarer color varieties, suggesting that their genetic
make-up be studied independently. These animals were
subsequently turned over to the writer for investigation.
During the winter two of them died without offspring
and hence only three were left the following spring.
Two of these proved to be what has been called by Castle
red-eyed agoutis and the third was a ‘‘ pink-and-red-
eyed’’ tortoise. The factors involved in the production
of these animals will be described below in more detail.

Castle (1914) had already described the different fac-
tors found in these animals and had given some indica-
tion of their relationships to other factors. At this
time he stated that red-eyed was allelomorphic to al-
binism, and that pink-eyed was recessive to the usual
dark-eyed condition. In 1915 Wright showed that red-
eyed formed an allelomorphic series with albinism and
dilute and intense pigmentation. This made the account
of the relationships of the factors fairly complete.

At about this time it occurred to the writer that with
the proper combination of the newly described factors
and other factors it should be possible to produce an ani-
mal which to all appearances would be an ordinary al-
bino, but entirely different genetically from what have
hitherto been known as albinos. Matings were imme-

1 Papers from the Department of Geneties, Wisconsin Agricultural Ex-

periment Station, No. 13. Published with the approval of the Director
of the Station.

120

.

No. 625] PINK-EYED SELF WHITE GUINEA-PIGS 121

diately started with this in view, but since there were
only three animals to work with at the start progress
was necessarily slow at first.

In the meantime Detlefsen (1916) had described some
pink-eyed white mice carrying the color factor. He be-
lieved they were due to the combination of the dominant
self white condition and the pink-eyed, but in his paper
this was not fully demonstrated.

Castle (1916) refers to two pink-eyed white guinea-
pigs, presumably albinos, which were born to pigmented
parents. The male parent was tested with true albinos,
but all of the eleven offspring obtained were pigmented.
Castle therefore supposes that the two pink-eyed white
offspring were not true albinos, but ‘‘ pink-and-red-eyed”’
animals lacking the factor for the extension of black or
chocolate pigment. This is undoubtedly the correct ex-
planation, as the evidence presented in this paper will
demonstrate. Castle, however, must have discarded his
pink-eyed whites before he realized their importance
since he makes no mention of testing them genetically.

Before going into further detail it may be well to de-
scribe briefly the various factors directly concerned in
the production of the pink-eyed self whites. Some of
those indirectly concerned are also described in order to
give the proper orientation.

DESCRIPTION OF FACTORS

B, the factor for black pigmentation. When unmodi-
fied by other factors black pigment is found only
in the eyelids, mammae, external genitals and the
.skin of the ears and feet.

b, the factor for chocolate (or brown) pigmentation.

E, the factor for extension of black or chocolate
throughout the pigmented part of the coat. The
unpigmented parts are of course white.

e, the partial extension factor. This factor causes
black or chocolate to be only partially extended
and therefore to appear in blotches, the remainder

Zii8.i 991866.2 £1830.)

Litter 1514 1202.4 2171.3

No. 625] PINK-EYED SELF WHITE GUINEA-PIGS 123

of the pigmented part being red. Guinea-pigs
spotted black and red in this manner are commonly
known as tortoises.

e, the non-extension of black or chocolate. In this case
black or chocolate are found only in the places men-
tioned under B, 7. e., in the eyelids, ete. E, e an
e form an allelomorphic series (Ibsen, 1916).
They are given in the order of their dominance.

C, intense pigmentation.

Ca, dilute pigmentation. Yellow is dilute red.

Cr, red-eyed or non-yellow. In a C, animal red (or
yellow) is never present in the coat and black or
chocolate are somewhat dilute. Owing to the ab-
sence of red the amount of pigment in the eyes is
decreased, thereby partially allowing the blood to
show through. It is because of this reddish tint to
the eyes that they have been called red-eyed by
Wright (1915). However, it seems more proper
to call them non-yellows since the absence of yellow
in the.coat is the more general and striking effect
produced by the factor. A C+ tortoise is there-
fore black-and-white because the red spots of the
ordinary tortoise are here absent. An eC animal
is a self red, but an animal with the composition
eC; is a self white, owing to the absence of red pig-
ment. Such an animal will be referred to in the
text as a ‘‘non-extended (e) non-yellow (C,).”’

From left to right: 911181, a gah gs non-extended non-yellow ;
esai an albino carrying both black (B) he extension factor (E);

g 1830.1, an na

to the pink color of the skin of the ears of Ọ 1118.1 agy g 1830.1, and also be-
ause of the a of the ai causing them cast shadows, the ears

appear darker than they really are. The pink eyes eleye appear dark.

Fig. 2. Gu seer g 849.1 is a bb dark-eyed non-extended non-yellow, ¢ 1258.1
isa a non-extended non-yellow carrying black (B).

Fig. 3. Guinea pigs 4 1202.1, a pink-eyed mon-extended non-yellow (ppC,0,ee),
9 1171.38, an ais heterozygous for partial extension (ep), and three of their
aoe litter 1514.

Fig. 4. A nearer and clearer view of g 1202.1 and Q 1171.8, already shown
in Fig, 3.

Fie. 5. Guinea-pigs 1202.1 and 9 1118.1, both pink-eyed non-extended n:
yellows ag one of their 24 offspring, § 1587.1, also a pink-eyed Demei
non-yello

124 THE AMERICAN NATURALIST [Vor. LIII

The C, factor has been described in some detail be-
cause it plays an important part in the production
of the synthetic pink-eyed self whites. C and Ca
are dominant to C+, while it in turn is dominant to

Ca:

Ca, albinism. Albino guinea-pigs generally tend to have
some pigmentation on the nose, ears and feet.
This varies in amount depending on the-other fac-
tors present. A be albino is almost if not quite
devoid of pigment (& 1830.1, Fig. 1), while a BE

‘albino is heavily pigmented at its extremities
(2 1866.2, Fig. 1). In all cases, however, the eyes
are pink and the greater part of the coat is white.

P, dark-eyed.

p, pink-eyed. The eyes are as pink as those found in
albinos. Black (B) or chocolate (b) are also af-
fected, becoming quite dilute, while red is unaf-
fected.

It will be seen from the description of the factors
given above that if we can obtain an animal which is a
non-extended non-yellow, eC,, and therefore a self white,
and which in addition is pp, or pink-eyed, we shall have a
pink-eyed self white (eC,p), which is not an albino in the
ordinary sense of the word as applied to guinea-pigs,
but which nevertheless to all appearances in an albino
(see 91118.1, Fig. 1). There is one difference, however.
Albinos vary considerably in the amount of pigment in
the ears, nose and feet depending on whether E, e°, or B
are present or absent. The synthetic pink-eyed self
whites on the other hand cannot carry E or e, but may
carry B. The pink-eyed factor (p) dilutes black pig-
ment (B) to such an extent that it is impossible to tell
by the appearance of the animal whether or not black is
present. All the synthetic pink-eyed self whites are
therefore a pure white and fulfil the guinea-pig fancier’s
requirements for a good albino better than most true
albinos do.

No. 625] PINK-EYED SELF WHITE GUINEA-PIGS 125

BREEDING OPERATIONS

As previously stated there were three animals with
which to begin breeding operations in the production of
the synthetic pink-eyed self whites. These were two
non-yellow agoutis, 360.1 and $361.1, and a pink-eyed
non-yellow tortoise, 2363.1. By various test matings
these were proven to be of the following gametic com-
positions: $360.1, AaPpC,C,e’e’,? 2361.1, AaPPC,C,ere?’,
and 2363.1, aappC,Cre’e’. It will be seen that all
three were homozygous for e”, the factor for the partial
extension of black or chocolate. On this account it would
be impossible by mating the animals together to produce
the pink-eyed self whites since the desired animals must
lack the extension factor and therefore be ee.

What seemed the best method of procedure was to
mate $363.1 (ppC,Cre’e’) to a self red (PPCCee), and
then to inbreed the F,’s (PpCC;e’e). These all looked
like ordinary tortoises. Since this was a tri-hybrid
cross and since the desired animals were to be homo-
zygous recessives, it would theoretically be necessary to
have 64 offspring for the production of one of the desired
type. This proved to be a very slow process and was
finally discarded in favor of other less methodical mat-
ings, which were more rapid in their results. The most
successful will be described in some detail.

A heterozygous tortoise $572.1, of the composition
PpCC,e’e obtained by mating ¢ 360.1 to a self red, was
mated to a yellow-and-white female, 629.3, carrying al-
binism (PPC.C.ee). One of their offspring, $849.1
(Fig. 2), was at first mistaken for an albino, but more
careful examination proved him to be a dark-eyed non-
extended non-yellow. The reason for mistaking him for
an albino was that he did not carry black (6).* His

2 A is the factor for agouti.

3 In a dark-eyed non-extended non-yellow (PeC;) the presence or absence
of black (B) is as easily detected as it is in an albino or a self red. In Fig.
2 gf 1258.1 carries black while ¢ 849.1 does not. In addition to having
much lighter colored extremities a bb non-extended non-yellow also has much
pinker eyes than one which carries black. For this reason it may be mis-
taken for an albino.

126 THE AMERICAN NATURALIST [Von LIH

gametic composition turned out to be PpC,Caee. He
was mated to a pink-eyed tortoise, 2? 734.2,4 whose com-
position judging by her offspring, must have been
ppCaCrere. This cross may be represented as follows:

g 849.1 x 2 734.2
PpC,Cgee ppCaCrere
E re plae
Gametes ae ne
pUre
pCae pCre
Offspring Obtained
PpCaC eve r
PpCaCaePe | Dark Syou tortoise i Me oa li aa 1
ppCaC ere C :
ppCaCaere } Pink prea? (OLLO = os aw aar aa A es 2
PpCaCree
PpCaCa } Dark-eved solt yellow oe S a Cle WN oe es 2
ppCaCree ie i
PP }Pink Protal- yelow oe int Vanes Goh ck EER 2 s
PpC,Cr :
PpC;-Caeve | Dark-eyed non-yellow tortoise .......:........... 2
ppC,C,ere i 4
pp0;0a } Pink eyed non-yellow tortoise .................. 0
PpC,C,ee
- - SVOHOW ar ei es 1
PpC,Caee | Dark eyed non-extended non-yellow
ppCrCree i r 2
ppC,Caee | Pint-eyea non-eatended non-yellow ........--.406
PORE oS ie a OO on a A ek 12

It will be seen that according to expectation there
should be equal numbers of 8 different phenotypes. The
12 offspring actually obtained are remarkably close to
expectation.

At the time: ihia cross was made the gametic composi-
tion of the mother, 2734.2, was not definitely known.
When, therefore, the two pink-eyed non-extended non-
yellows were born one could not be certain that they
were not just ordinary albinos. The only definite method

4 Ancestry of 9 734.2:

8 179.1, PPCCaee
Q 734.2, ppCaC ere 2 363.1, ppC,C,erer
|9 549.3, PpCrCrever S sere? pee oe

g 346.1, PpCaC;ere

No. 625] PINK-EYED SELF WHITE GUINEA-PIGS 127

to test this was to mate them to true albinos, and this was
accordingly done.

One of these animals, 21118.1 (Figs. 1 and 5), mated
to an albino, $596.3 (PPC.Cu.ee), had 4 dark-eyed non-
extended non-yellow offspring and 2 albinos. This would
make her composition ppC,Ccee and the cross may be
represented as follows:

g 596.3 x © Q181
PPC,Cagee ppCrCaee
pC re

Gametes: 3 PC,e

Offspring: PpC,Caee, dark-eyed non-extended non-yellow;
PpCaCaee, albino.

The other animal, (1202.1 (Figs. 3, 4 and 5), was
mated to an albino (Ẹ 1171.3, Figs. 3 and 4) of the cvm-
position PPC.C.e’e. There were 12 offspring. Of these
6 were dark-eyed non-yellow tortoises and the other 6
were dark-eyed non-extended non-yellows (see litter 1514,
Fig. 3). This would indicate that ¢1202.1 was of the
composition ppC,C,ee, and the cross would be as follows:

1202.1 x ọ 1171.3
ppCrOree PPCaCae?e
Gametes: $ pore ey
Offspring: PpC,C,ere, dark-eyed non-yellow tortoise ;6
PpC,C,ee, dark-eyed non-extended non-yellow.

It will thus be seen that the two pink-eyed non-ex-
tended non-yellows were not of the same composition,
the male, 1202.1 being homozygous for C,, while the
female, 1118.1, was CCa. This is what one might expect
by reason of their parentage.

According to expectation, these pink-eyed non-ex-
tended non-yellows when bred together should have

5 The albinos used in these test matings were unquestionably homozygous
dark-eyed (PP) since pead came from stock that has never been known to
T w pink-eyed factor (p)

e dark-eyed non-yellow tortoises as sa as the dark-eyed non-ex-
rs non-yellows are being inbred, and the results from these matings will
be reported at some future date.

128 THE AMERICAN NATURALIST (Von. LII

nothing but pink-eyed non-extended non-yellow offspring,
Thus:

4 1202.1 x Q 1118.1
; ppCrCree ppCrCaee
af pCve
Gametes: 1 pcre a pe pte
Offspring: 4 if ccxeet pink-eyed non-extended non-yellows.

The mating has been made and thus far there have been
24 offspring, all of them pink-eyed non-extended non-
yellows (see Fig. 4). Some of these offspring are being
tested by being mated to albinos, but as yet their compo-
sition with respect to the presence of albinism is not
definitely known. It may be of interest to mention that
one of them when mated to a PPEE albino had one dark-
eyed self black offspring. The other two offspring in
the same litter were black-and-white, due to the fact that
neither parent was homozygous for entire pigmentation

Discussion

In most domesticated mammals an albino, as ordi-
narily understood by breeders, is a completely self white
animal with pink eyes. Albino rabbits are of this type.
In addition there is the Himalayan variety which also
has pink eyes, but the coat instead of being entirely white
is pigmented at the animal’s extremities. This condi-
tion is dominant to albinism and recessive to the fully
pigmented condition, thus forming part of an allelo-
morphic series (Sturtevant, 1913). Guinea-pig fanciers
also have what they call a Himalayan variety. Here,
however, the genetic relationship differs from that found
in rabbits. Himalayan guinea-pigs are undoubtedly true
albinos carrying the factor for black (B) and the exten-

7 An § animal is entirely pigmented, while one that is ss shows some white
spotting and is therefore not entirely pigmented. In this connection it may
be of interest to note that in a non-yellow tortoise the white spotting may be
due to two different causes, (1) because of the non-yellow factor (Cp) the
yellow of the ordinary tortoise is here white, and (2) if the animal is ss it
will show some white spotting on this account also,

No. 625] PINK-EYED SELF WHITE GUINEA-PIGS 129

sion factor (Æ) which extends the black. The presence
of these two factors tends to make the extremities heay-
ily pigmented.

The albino guinea-pig, according to the fanciers’
standard, should be as completely white as the albinos
of other species. This, however, has never been entirely
attained, even though much selection has been practised.
Albinos most nearly approaching the standard lack both
black and extension factor and are therefore eb. When
mature they ordinarily show, nevertheless, a lightly pig-
mented rim along the edges of the ears. The synthetic
pink-eyed whites, produced as explained in the earlier
part of this paper, are, on the other hand, pure white,
and they therefore satisfy the fanciers’ standard in this
respect. They would meet his desires, furthermore, in
that they breed true for this character.

There are other means besides those already men-
tioned whereby pink-eyed self white guinea-pigs could be
produced which would satisfy the fancier’s standard.
One method would be to combine a self white condition
described by Castle (1905) with the pink-eyed. Self
whites of this type, however, do not breed true. Very
frequently they throw spotted offspring. They seem to
be merely an extreme form of white spotting, all of the
animal being unpigmented except the eyes. Pink-eyed
self-whites of this type would necessarily on this ac-
count be very unstable in the transmission of their coat
character.

Another method would be to prodyce a ‘‘ pink-eyed ”’
(pp) albino. The pink-eyed factor in this case would
cause the pigmentation ordinarily found in albinos to
become invisible. We have produced an animal of this
type. When mated to pink-eyed it had nothing but pink-
eyed offspring, and when mated to albinos it had only
albino offspring. It was pure white in color.

It seems quite probable that synthetic pink-eyed self
whites may also be produced in rats. Castle (1914) has
already described yellow varieties and pink-eyed varie-

130 THE AMERICAN NATURALIST [Von. LIII

ties, and Whiting (1916) has stated that the non-yellow
factor has been found in some animals.

In conclusion it may be said that we have furnished
one more proof of the fact that the phenotypic appear-
ance of an animal may entirely mislead one as to its
gametic composition. The synthetic pink-eyed self white
guinea-pig may also serve another purpose. By proper
matings an animal can be produced which carries all the
known recessive color factors in guinea-pigs except al-
binism, and animals of this type should be most useful
in determining the possible linkage relations between
the factors.

LITERATURE CITED
Castle, W. E. ‘
1905. Heredity of coat characters in guinea-pigs and rabbits. Carnegie
Institution of Washington, Publication No. 23, 78 pp. 5 pls.
1914. Some new varieties of rats and guinea-pigs and their relation to
oblems of color inheritance. AMER. NAT., Vol. 48, pp. 65-73.
Castle, W. E., ghy Sewall Wright.
1916. Studies of inheritance in guinea-pigs and rats. Carnegie Insti-
tution of Washington, Publication No. 241, 192 pp.
Detlefsen, J. P
1916. Pink-eyed white — carrying the color factor. AMER. NAT.,
Vol. 50, pp. 46-4!
Ibsen, H. L
1916. Tricolor inheritance. I. The tricolor series in guinea-pigs.
etics, Vol. 1, pp. 287-309.
Sturtevant, A. H
1913. The Windleyss rabbit case with some eT ee on es
allelomorphs. AMER. Nat., Vol. 47, pp. 23
Whiting, P. W.
1916. A new color variety of the Norway rat. Science, N. S., Vol. 43,
p- 781,

PARTHENOGENESIS AND CROSSING-OVER IN
THE GROUSE LOCUST APOTETTIX!

PROFESSOR ROBERT K. NABOURS

Kansas STATE AGRICULTURAL COLLEGE AND EXPERIMENT STATION

INTRODUCTION

Tere have been found among the grouse locusts, genus
Paratettix, in nature, fourteen factors for color pat-
terns, all in one series of multiple allelomorphs. A fif-
teenth factor, a modified S, in the same series, has orig-
inated in the laboratory by means as yet not understood.
Another factor, 9, for melanism, discovered in nature,
has been found to segregate independently of the multiple
allelomorph group (Nabours, 714, °17). Still another
factor, $, for red-all-over, yet to be described, also from
nature, behaving precisely as does @, though segregating
independently of both it and the multiple allelomorph
series, has been bred in considerable numbers. In one
species of the genus Tettigidea,’ bred in our laboratory,
there have been described a series of five multiple allelo-
morphic factors for patterns, and an independently seg-
regating color factor of the behavior of 4, or ¢, in Para-
tettix (Bellamy, 717). In these experiments, involving
several hundred kinds of matings and many thousands
of individuals, with only the two exceptions, both as yet
unexplained, segregation has taken place as expected.

Among another genus, A potettiz,* of the grouse locusts

1 Paper 25 from the "gaia Laboratory, Kansas Agricultural College
and Experiment Statio

2 Paratettix texanus Hancock Kindly identified ee both Dr. J. a Han-
cock and Mr. Jas. A. G. R

3 Tettigidea parvipennis allie Hancock.

4 Identified by Mr. Rehn as Apotettix eurycephalus Hancock, and by
Doctor Hancock as follows: ‘‘nearer the Mexican species Apotettix con-
verus Morse, than the nearly allied Texan species, Apotettix eurycephalus
Hancock. Inasmuch as you have used material from both Texas and Mexico
in your experiments, it is possible you have hybridized the two.’’ The nat-
ural history of this group has been described (Hancock ’02).

131

132 THE AMERICAN NATURALIST [Vou. LII

there have been discovered in nature eleven factors for
color patterns, all in the same series, but evidently only
a few, if any, are allelomorphs. Pending further con-
sideration, the patterns are designated as AA, GG, KK,
MM, OO, RR, TT, WW, XX, YY and ZZ, respectively
(all conspicuous, except AA which is of a mottled gray
ground color and well protected). These patterns are as
sharply defined and distinct, each from any other, as are
those of Paratettix. Any two make a readily recog-
nizable hybrid pattern, with the elements of each parent
pattern seemingly equally represented, except that the
part of a hybrid pattern produced by the factor A, when
it is a member, is less clearly perceived, and such hybrids
can not be, in every case, superficially distinguished from
the pure, or homozygous, pattern of the more apparent
member. For instance, it requires trained and careful
scrutiny to distinguish between AK and KK, AW and
WW, AY and YY, etc. It appears that the pattern AA,
if it be the result of only the one factor, is quite different
from the others, not only in its manifestation in the
hybrid of which it may be a member, but also, as will be
shown later, in that it appears to result from crossing-
over among the others. On the other hand, KW, KY,
WY, and most other hybrid patterns not containing A as
a member are sharply distinct, each making a composite
picture of both components. Even in ease of linkage,
where three, four or more factors are combined, the in-
dividual presents the composite appearance of all the
patterns involved. For example, in KMR, KMR or
KRM, KYZ and KRYZ individuals all the patterns con-
cerned are clearly visible and apparently equally repre-
sented. In respect to representation of patterns in the
hybrids, as well as in the actual resemblance of a few
of the patterns, there is striking parallelism between
some of the members of this genus and some of those of
Paratettix.
However, as already suggested, in contrast with the
inheritance behavior in Paratettix, most of the factors

No. 625] PARTHENOGENESIS AND CROSSING-OVER 133

in A potettix show from small to considerable percentages
of crossing-over. The discovery of crossing-over, with
the further disclosure that these forms breed partheno-
genetically, as well as bisexually, prompts me to submit
a preliminary report in advance of a more extended
presentation of the data and illustrations which can not
be made ready this year.

The Adams fund has cared for the expenses, and I
have had the generous and open-minded support of di-
rector, now president, W. M. Jardine. Mr. A. W. Bel-
lamy gave effectual assistance during the earlier stages
of the experiment.

PARTHENOGENESIS AND CROSSING-OVER IN THE FEMALE

The discovery of parthenogenesis in Apotettix was
the result of attempts at cross-breeding the members of
this genus with those of the genus Paratettix. It was
observed that when an Apotettix male was used with a
Paratettix female no progeny ever resulted, but the
female of the reciprocal gave offspring, exclusively
females, and of her own color pattern if she were homo-
zygous, or segregated into her components and cross-
overs if she were heterozygous. Then it was soon ascer-
tained (see chart) that the Apotettix females which had
never been exposed to males of any kind at any time be-
haved in this respect precisely as did those exposed to
Paratettix males. Copulation between members of the
two genera was never observed. Confirming these ob-
servations, Dr. J. L. Hancock kindly examined speci-
mens for me and concluded that on account of structural
differences the members of the one group could not mate
with those of the other.

The chart illustrates a portion of the experiment which
definitely determined that parthenogenesis occurs. It
also shows crossing-over in the females. Following the
chart, the first KK female, exposed to a Paratettix male
of strongly contrasting pattern, produced 10 offspring,
all females and of her own pattern. Three of these F,

134 THE AMERICAN NATURALIST [Vou. LIII

Paratettix x KK

pole
o-i
E T
Ptx x KK Ptx x kK Ptx x KB
i te air
Same9 0-7 0-10 OIA
| : | | EKK l SKK | | KK
KK? RR x KK TT x KK Ptx x KK=0-7 Ptxx KK=0-4 RY x KK Ptx x KK=0-1
on tke | | tx x KK=0-8 Ptx x KK=0-49 Ptx x KK=0-2
KK KR KT KK Ptx x KK=0-12 Ptx x KK=0-4 ER KY Q Ptx x KK=0-8
0-1 49-39 10-4 0-15 0-27 Ptx x KK=0-7 7-12 9-13 0-32 Ptx x KK=0-9
a w w UH A > 2 =0-68 ST Ty ‘Se Ptx x KK=0-22
88 14 15 ET L x] Re=0-41 à x KE=0-22
0-173 0-64
Y 173 ‘ 64
Same? a.
Te = EM x KK KK9? KK9? KY99 KK9¢
sf Sete DENEN Pen: EA P RUR | aA +

Ky kk kk kM a: ee E oe
20-28 Q-10 7-6 13-19 0-40 Q-163 0-1 0-1 0-2 0-0 0-
48 10 15 25 3-14 165 9-0 0-11 0-8 0- 4.

as Q3 Q12 Q5 = Q-59

2 2B e

individuals were separately exposed to Paratettix males
and gave 7, 10 and 14, respectively, and again all were
KK females like the female parents and grandparent.
One of these F, females was subsequently mated to a
male of the strongly contrasting pattern MM, of her own
genus, and then she produced 12 males and 16 females of
the composite pattern KM.

From left to right, in F,, a female KK was mated to
an RR male and gave, in F}, 4955:3999 of the inter-
mediate KR pattern. About three weeks after the death
of the male this KK female was removed to another cage,
where she produced a few offspring, the only one re-
corded being a KK female. The next F, KK female was
mated to a TT male of a strongly contrasting pattern.
She produced 1044:499 of the hybrid pattern KT, and
15 KK’s, all females and without a trace of the TT pat-
tern. The following three F, KK females were placed
with Paratettix males and produced 7, 8 and 12 KK
females, respectively. The next four F, KK females
were also exposed to Paratettix males, and gave 4, 49, 4
and 7 offspring, all like the female parents; while the re-
maining two sisters, without exposure to males of any
kind at any time, gave, in F, 68 and 41 KK females,
respectively. :

One of these, the one having produced 68 offspring

No. 625] PARTHENOGENESIS AND CROSSING-OVER 135

Same 9
pe
pe TEN
KM
12-16 F:
MT x KK OY x KK RR x Ptx x KM=0-0-0 0-2 0-5 0-0
aee ieoi Ee ee
: x x KM=0-0- 0-0
Q-4 11-21 13-26 13-4 9-17 0-0 —x 2KM=0-0-1 0-15 0-18 0-1
4 32 39 14-22 12-18 1-1 0-0-} 0-26 0-36 0=]
sik tai 27-26 21-35 1-1 26 Z8
53 56 2
Explanation.
KK? MM x KK YY x KK Ptx=Paratettix male. Males on the left;
aE a ass aR females on the right be a 27°
KK KM Ky KK females. O0-0O-1=Sex not dete
2-2 0-3 0-1 O-3 Fa 49-39-49 males:39 females. All other
eer r ae ee adipi refer to factors in Apotettix anā
pattern

parthenogenetically, was mated to a YY male and
gave KY 2044:2899 and 10 KK 22 without a trace of
the YY pattern. The other F, KK female, which had
given 41 offspring parthenogenetically, when mated to a
KM male, gave KK 745:6992, and KM 1355:1099. It
is obviously impossible to determine if all the KK off-
spring from this mating were produced bisexually, or
some of them parthenogenetically. However, since par-
thenogenetically produced individuals are, with rare ex-
ceptions, females, the result is somewhat checked by the
proportion of 75:699. From the 68 F, offspring sev-
eral females were allowed to reproduce without exposure
to males of any kind at any time, and 163 KK females
resulted in F,. From the 41 KK F, individuals one
which had never associated with any male gave 40 - KK
females in F,, and another, also parthenogenetically,
gave 14 females and 1 male, all KK’s.

The following F, KK female was mated with an RY
male and gave KR 75:1299, KY 95:1399, and 32 9?
of the pattern of the KK parent. Two of these F, KK
females were further tested parthenogenetically and pro-
duced, in F,, a total of 50 offspring, all like themselves.
Two of the KY 99 were bred parthenogenetically and
gave, both combined, 1 AA:12KK:10YY: 2KYKY, all
females, thus exhibiting the crossovers AA and KYKY,
as well as the expected segregates KK and YY.

136 THE AMERICAN NATURALIST [Vou. LIH

The five next F, KK females placed with Paratettix
males produced 1, 2, 8, 9 and 22 offspring, respectively,
all KK females. The sixth female of this group, ex-
posed to no male whatsoever, behaved precisely as did
those individuals which had been with Paratettix males.
Another of these F, KK females was placed with an
MT male with 4 KK female offspring resulting in
F, none evidencing any M or T part in the par-
entage. Three of these were bred further and gave off-
spring in F, as follows: one, without any male, 2 KK
females; the second, mated to an MM male, 3 KM females
showing the male parentage unmistakably; and the third,
mated to a YY male, 1 KY female showing male par-
entage clearly, and 3 KK females indicating none of the
male characteristics and plainly parthenogenetic. On
account of the extremely small numbers involved, I sug-
gest there is no special significance to the absence of
males in the two last described matings. An eighth F,
KK female, of this group, was mated to an OY male and
gave KO 1155:2199 and KY 1365: 26 99.

The following two matings show RR males mated to
KM females and giving KR 27 64:26 992, MR 214: 35 99
and the crossovers KMR 16:19. Three of the KM
females were placed with Paratettix males, but the off-
spring exhibited no more evidence of male parentage
than those produced by the two KM sisters without ex-
posure to any male. The combined result from the five
KM sisters was AA 1:KK 26:MM 38: KMKM 1, all fe-
males, except that the sex of the AA individual was not
determined. As in the ease of the KY females already
noted, this furnishes a very interesting exhibition of
segregation as well as crossing-over in parthenogenetic
individuals.

Subsequently from KK females, individually and in
groups, but not exposed to males of any kind at any time,
there have been given 2,726 female and 4 male offspring,
all KKs, some of them having arrived at the fifth par-
thenogenetic generation. Including the KK females rep-

No. 625] PARTHENOGENESIS AND CROSSING-OVER tor

resented in the chart, and others exposed to males with-
out effect, there have been produced parthenogenetically
from KK females, a total of 3,289 females and 5 males,
all of the KK pattern. Other females than KK (some
of the data presented below) have produced 1,181 fe-
males and 2 males, making a total of 4,470 females and
7 males of various patterns produced parthenogenetically
(August 1, 1918). Individuals of all the patterns, except
AA which has not been adequately tested, have given off-
spring by parthenogenesis. It is not known if any one
breeds in this way more readily than any other, the KK’s
having been used up to the present time more than the
rest. Furthermore, it has not been determined whether,
or not the capacity for parthenogenesis is in any sense
an inheritable character. There are, however, great dif-
ferences among individuals in this respect, even from
the same F,, or F,, parthenogenetic batch.

Note: ġġ on the left.
99 on the right.

mo 2g RR? 2 eg 29 ag 992 uren Q
MM RR YY Yuyz MRMR
0-1 0-1 0-2 0-18 0-1
0-1 0-16 0-4 0-7 0-16
0-7 0-61 0-5 0-19 Total 0-17
0-12 0-7 0-22 Total 0-44

0-15 0-24 0-8

0-37 0-85 Total 0-41

0-2
Total 0-96 Total 2-168

(6) (7)

ER FẸ KY? F
AA KK- MM KMKM AA KK Y¥ KYKY
18 oi 08 pA 0-0 06 0-2 00
0-0-0 05 O05 00 6) 68 68 0-0
0-0-0 0-13 0-12 0-0 0-0 -01i 00 oO
Total 0-0-0 0-19 0-20 OI oo 04 08 00
From Chart 0-0-1 0-26 0-88 0-1 Total 0-0 0-14 0-14 0-0

Total 0-0-1 0-45 0-58 0-2 From Chart 0-1 0-12 0-10 0-2
Total 0-1 0-26 0-24 0-2

138 THE AMERICAN NATURALIST @ [VoL. LIH

(8) (9)
KYZ 29 ma Q
KK Yuvu KYKY zz Ax MM RR MRMR
0-3 0-11 0—0 0-1 0-2 0-11 0—4 0-2

Total 0-21 0-25 0-2 0-8

A few other results from breeding Apotettix females
parthenogenetically are given in the above tables, which
also include the segregation and crossing-over in the KY
and KM females shown in the chart.

Further crossing-over is indicated in the partheno-
genetic KYZ females (8) and the MR female (9). The
latter indicates more than 21 per cent. of crossing-over,
but the total crossing-over shown in 296 offspring from
MR females, produced both bisexually and partheno-
genetically, amounts to only 20, or less than 7 per cent.

few other simple hybrid females, some partheno-
genetically and others bisexually, have produced cross-
overs as follows: |
From GM females 279 individuals with 11 crossovers, about 4 per cent.
From KM females 517 individuals with 5 crossovers, about 1 per cent.
Y females 205 individuals with 12 crossovers, about 6 per cent.
From RY females 33 individuals with 3 crossovers, about 10 per cent.
From TY females 70 individuals with 8 crossovers, about 12 per cent.
From RT females 125 individuals with no crossovers.

There is every indication that as the numbers avail-
able become larger these percentage figures will be dif-
ferent; therefore it seems inadvisable to project at this
time even a tentative diagram illustrating crossing-over
percentages.

CROSSING-OVER IN THE MALE

Crossing-over in the females, in parthenogenetic as
well as bisexual reproduction, is shown in the chart, and
tables (1-9), and there are numerous other cases to be
presented later in both bisexual and parthenogenetic
breeding. While it appears that the crossovers in the
females greatly exceed those in the males, the data are

No. 625] P4RTHENOGENESIS AND CROSSING-OVER 139

as yet insufficient to justify a final judgment. A few of
the considerable number of cases of crossing-over in
males are herewith given:

0) ap (12)
AYE BAK Ore ee Ma x Ox
AK AYZ KYZ AZ KO KTY KT + GM GR KM KR KMR
3-2 3-2. 2-3 6-1 0-1 9-12 11-17 2-1 14 35 1-5 12 0-1
oer ig) net: oes ae es aes eae eae oe Woes Bicone iB
a3) a4)
KM x RR KYM X RT

ÉMRARRR MR MT KYR KYT KMT

6-7-5 3- 44 0-1 0-2 8-15 14-9 16-20 144 0-1

SMe See 8, a Se 23 opps | Wee | gees

The KM x RR mating (13) is of interest and a sample
of a frequent occurrence. The KMR individual, of un-
mistakable pattern, could not be accounted for otherwise
than by assuming crossing-over in the male, but since AR
and RR were so much alike and were not bred further,
we can not know whether both of these were AR, the A
gamete coming from crossing-over in the male, or whether
they were RR produced parthenogenetically by the RR
parent, or one was produced by the former and the other
by the latter method.

Discussion’

Are the female gametes in this group of grouse locusts
all of the same kind with respect to the necessity of fer-
tilization, or do some of them require the spermatozoon,
in order to develop, and others not? The latter situation
is suggested by the fact that mated individuals frequently
reproduce bisexually and parthenogenetically at the
Same time. Also, often, when an individual which has
been reproducing parthenogenetically is mated she
thenceforth gives offspring, some exhibiting, and others
not, male parentage. In an unmated female perhaps the
eggs that require fertilization disintegrate either before
or after oviposition. (The eggs are oviposited in the

140 THE AMERICAN NATURALIST [ Vou. LIII

ground. This matter can be, and is being, investigated.)
On the other hand, it as often happens that when an in-
dividual reproducing parthenogenetically is mated she
thenceforth gives offspring all showing male parentage.
Also, though less readily, a mated female reproducing
exclusively bisexually, when placed to herself, will, after
a few weeks, give offspring parthenogenetically. Al-
though the end result data which might give light on this
point have not as yet been adequately developed, I ven-
ture the suggestion that with respect to the need of fer-
tilization the mature eggs are approximately the same,
and that it is the time of the entrance of the sperma-
tozoon which determines the matter. If a spermatozoon
enters the egg at the proper stage of its maturation the
pronuclei unite; if no spermatozoon enters, or one enters
too late, the egg either proceeds parthenogenetically, or
fails to develop altogether.

The diploid number of chromosomes in the A potettix
female appears to be fourteen. This number has been
clearly demonstrated in the late metaphase plate of an
oogonial division in an individual derived bisexually. In
a preparation of somatic cells of a female produced par-
thenogenetically, nine apparently whole chromosomes
and some fragments were observed in one (Mr. A. H.
Hersh, unpublished.) The females of some other mem-
bers of the Tettigide have fourteen chromosomes (Rob-
ertson, 716). In some forms of the Tettigide the males
have been shown to have thirteen chromosomes as the
complete number (Harman, 715, and Robertson, 716).

The ‘‘maneuvers of the chromosomes’’ theory of Mor-
gan may very well account for the observed end results
presented in this paper, though the possibility of some
other explanation is not by any means excluded. Even
if the results herein entitled ‘‘crossing-over’’ should at
some time be found actually not to be connected with
the maneuvers of the chromosomes, the term might still
be retained as an adequate expression of whatever does
occur. The discussion of the mechanism, physiological

No. 625] PARTHENOGENESIS AND CROSSING-OVER 141

processes, or both, involved in the parthenogenesis must
await further investigation.

In nature, individuals of the pattern AA, of a mottled
gray ground color, in striking contrast with, and much
less conspicuous than, the rest, exceed in numbers
all the others combined. This situation has been ac-
counted for in the past by the assumption that the bril-
liant patterns rendered the individuals possessing them
so conspicuous that they more readily fell prey to en-
emies, while the inconspicuous and protected AA indi-
viduals were largely unmolested. Now, since it has been
demonstrated that crossing-over among the forms of con-
spicuous patterns produces the A gametes as well, the
cause of the preponderance of AA patterns in nature
may, in part at least, call for an entirely different ex-
planation. AA may be the primitive form and the others
have originated from it, by mutation or in some other
way. The form AA seems to correspond to the so-called
‘normal’? or ‘‘ wild type,’’ though all the others have also
been found exclusively in nature, none (in Apotettix)
having so far originated in the laboratory.

CoNCLUSIONS

1. Through complete isolation of females from males
of any kind, in some cases for as many as three genera-
tions, and, in addition, by genetic behavior, it has been
demonstrated that these forms of Apotettix are gyno-
genetically, except rarely parthenogenetic (tychoparthe-
nogenetic).

2. Segregation is demonstrated as occurring in hetero-
zygous individuals reproducing by parthenogenesis, as
well, and apparently to the same extent, as in those
females reproducing bisexually.

3. Crossing-over is demonstrated as occurring in het-
erozygous individuals reproducing by parthenogenesis,
as well, and apparently to the same extent, as in those
females reproducing bisexually.

4. Crossing-over occurs in the male, as well, but ap-
parently not to the same extent, as in the female.

142 THE AMERICAN NATURALIST [Von. LIII

LITERATURE CITED

re A. W. 1917. Studies of Inheritance pe Evolution in Orthop-

i Journal si aea Vol. VII, No.
RS J. 1902. e Tettigidae of N. A. iila
-m eni F: 183. Spermatogenesis in Paratettix. Biol. Bulletin,
Vol. XXIX, No.
eo T. H and TIR 1915. Mechanism of Mendelian Heredity.
New Yor

parae Robert K. 1914. Studies of Inheritance and Evolution in Orth-

Journal of Genetics, Vol.
1917. Stu

y f: LEL No:8,
: Seas Robert K, Inheritance and eget in Orth-

optera IT and ont Journal of Nie VOL VIL NO: L
Robertson, W. R. 1916. Chromosome ‘Studies I. Jour. Morphology,
Vol. XXVII, = 2.

THE EVOLUTION OF ARTHROPODS AND THEIR
RELATIVES WITH ESPECIAL REFERENCE
TO INSECTS!

G. C. CRAMPTON, Px.D.

MASSACHUSETTS AGRICULTURAL COLLEGE

THe two lines of descent which have culminated in the
production of some of the most active and ‘‘dynamic”’
of living creatures, and those in which the psychic facul-
ties have attained their highest degree of perfection, are
represented by the vertebrate group Mammalia, at whose
apex is man, and the invertebrate group Arthropoda, at
whose apex are the Insecta. Since these are the great
rival groups, contending for the possession of the earth,
the tracing of the paths by which they have arrived at
their present dominating positions affords one of the
most fascinating chapters of the study of organic evolu-
tion. Concerning the ancestry of man, there is some
degree of agreement of opinion in modern works upon
the subject; but the recent investigations of Handlirsch, |
1904-1908, are not in accord with those of his predeces-
sors in the study of insect phylogeny, and since his views
have received a surprisingly widespread acceptance, his
work has served to revive the discussion of the ultimate
derivation of the insectan type of animals.

There can be but little doubt that the insects and their
arthropodan relatives are the descendants of ancestors
related to the worm-like forms. These in turn were de-
rived from lower worm-like ancestors resembling the
members of the Rotifera-Platyhelminthes group. In the
present state of our knowledge of the subject, it is hardly
possible to state with any degree of certainty whether
the ancestors of the worm-like forms were ultimately

1 Contribution from the ORE Laboratory of the Massachusetts
Agricultural College, Amherst, Mass

143

144 THE AMERICAN NATURALIST (Vou. LIII

derived from the Celenterata-Porifera group—i. e., from
a celenterate (cnidarian ?) type of animal through
Ctenophora-like (?) forms; or more directly from a
colonial protozoan type through forms comparable to the
‘‘Mesozoa’’ and their relatives, such as Dicyema, ete.;
although there is a strong probability that the lower
worm-like forms arose from ancestors occupying a posi-
tion intermediate between these types of animals. From
the standpoint of evolution, the Rotifera and Platyhel-
minthes (also such worms as Dinophilus, ete.) are among
the most important of the lower worm-like forms, since
they have departed as little as any from the condition
characteristic of the ancestors of the ‘‘Annelida’’ and
many other invertebrates; and even the line of descent
of the vertebrates themselves may ultimately lead back
to forms not unlike the members of this group. ‘A higher
stage of development is represented by the ‘‘ Annelida”?
(including the Sternaspidide, Gephyrea, etc.), which are
a group of the greatest phylogenetic importance due to
the fact that their line of development is approached by,
or is paralleled by, those of many other invertebrate
groups, and to the fact that they have retained a great
number of characteristics apparently typical of the an-
cestors of the Arthropoda. Their forebears probably
occupied a position intermediate between the Rotifera
and the Platyhelminthes, and indeed, some investigators
have even placed the ‘‘archiannelid’’ Dinophilus among
the planarian Platyhelminthes, although its closest affini-
ties seem to be with the annelidan worms Protodrilus and
Polygordius.

From their annelid-like forebears, there have branched
off two important lines of descent, which have ap-
proached very close to the arthropodan type, and which
have even been classed among the Arthropoda by some
investigators. One of these lines of development is rep-
resented by the Onychophora, which are suggestively
arthropod-like in many particulars, although they have
retained many annelidan characters; while the other line

No. 625] THE EVOLUTION OF ARTHROPODS 145

is represented by the Myzostomida, which are regarded
by some authorities as occupying a position intermediate
between the chetopod Annelida and the Tardigrada.

The position of the Tardigrada is still a subject of con-
siderable discussion, and the decision of the matter de-
pends largely upon the settling of the question whether
the apparent simplicity of their organization is due to
the retention of a primitively simple condition, or to a
secondarily acquired simplicity brought about by reduc-
tion or degeneracy, ete. If the simplicity of the Tardi-
grada is a primitive one, there is much to be said in favor
of placing them next to the Myzostomida in the Myzo-
stomida-Onychophora group; but whether the Lingua-
tulida should also be included in this group seems very
doubtful.

From the resemblance of the adults to Eriophyid mites,
and of the immature forms to such short-bodied mites as
Phytoptus, ete., many recent authorities would place the
Linguatulida near the Acarina—a highly modified group
far removed from the base of the arthropodan stem. If
this view is correct, the apparent simplicity of the Lin-
guatulida is to be regarded as the result of a reduction or
degeneration rather than the retention of a primitive
condition, in forms so far removed. from the base of the
arthropodan stem; and if the simplicity of the Tardigrada
is primitive, while that of the Linguatulida has been sec-
ondarily acquired through reduction or degeneracy, the
apparent resemblance between the Tardigrada and
Linguatulida must be regarded as the result of conver-
gence, or parallelism, rather than of consanguinity. `
Under these conditions the Linguatulida could not be
grouped with the Tardigrada, if the latter are placed
next to the Myzostomida in the Myzostomida-Onycho-
phora group; but if the apparent simplicity of the Tardi-
grada is also due to reduction or degeneracy, they too
might be placed with the Linguatulida near the mites—
although this does not appear to be very probable from
our present knowledge of the subject. It is quite ap-

146 THE AMERICAN NATURALIST [ Von. LII

parent, however, that the matter needs considerable
further investigation before this question can be definitely
determined.

The affinities of the molluscan group are somewhat
obscure, but the study of their immature stages would
indicate that the Mollusea are rather distantly related
to the Chetopoda, Gephyrea, ‘‘Polyzoa,’’ and other
annelid-like forms. If this be true, their line of develop-
ment apparently likewise extends back to forebears simi-
lar to the members of the Rotifera-Platyhelminthes
group (which are very like the ancestors of the ‘‘ Anne-
lida’’ also), and the ancestors of the Echinodermata and
Hemichordata may possibly be traced back to a similar
stock (or to forms closely related to them) as well.

The Hemichordata are regarded by many authorities
as a group which has departed but little structurally
from the ancestral condition of the forms leading up to
the vertebrate type of animal. From a study of their
larve, some investigators have concluded that the Hemi-
chordata are related to the Echinodermata; but this
brings us no nearer to the solution of the problem, since
the Echinodermata occupy an isolated position, and their
ultimate affinities are very obscure. Although the Echi-
noderm line of development may lead back more directly
to the Coelenterata, there is a possibility of their fore-
bears being quite closely related to the members of the
Rotifera-Platyhelminthes group which have departed
but little from the condition characteristic of the ances-
tors of the ‘‘ Annelida ’’ and other invertebrate groups;
-and even if the Echinodermata are to be regarded as the
descendants of Celenterata-like forebears, it must be re-
membered that the ancestors of the Rotifera-Platyhel-
minthes group were themselves very closely related to
the Coelenterata, and would probably have been classed
as such, rather than as colonial Protozoa.

It is quite possible to regard the line of development
of the Echinodermata as related to the other two lines of
development in question, and as branching off near, or

No. 625] . THE EVOLUTION OF ARTHROPODS 147

even further down in the developmental scale, than the
points of origin of the lines of descent of the Hemichor-
data and the worm-like forms. It is also possible to
suppose that in the hemichordatan line there have been
carried over certain developmental tendencies from their
common ancestry, such as the preservation of a ‘‘tor-
naria’’ larva characteristic of the Echinodermata, while
in other respects the Hemichordatan line of development
has paralleled that of the worm-like forms more closely,
having taken over more of the tendencies which were to
find opportunities for fuller expression in the worm-like
forms from their ultimately common inheritance. Under
these conditions the Hemichordata are related to both
the Echinodermata and the worm-like forms, but their
line of development has accompanied that of the worm-
like forms much more closely and for a longer distance
before each branched off along its own path of speciali-
zation. If the lines of development of the Hemichordata
and of the worm-like forms have an ultimately common
ancestry, and if both lines of descent have ‘‘travelled
along the same developmental road’’ for a considerable
distance before each branched off along its own path of
specialization, it is not surprising that we find many
structural resemblances in the two lines of descent, and
the resemblance of such Hemichordata as Cephalodiscus,
Rhabdopleura, Phoronis, etc., to certain ‘‘ Polyzoa,?’ may
be as much the result of consanguinity as of ‘‘convergent
development.’’ This view enables us to harmonize the
apparently discordant theories concerning the ultimate
origin of the Vertebrata—all of which may contain a por-
tion of the truth, as is frequently the case in the different
hypotheses put forward to explain certain observed phe-
nomena. Thus, according to this conception, we may
derive the Vertebrata from forms similar to the Hemi-
chordata, and still account for the annelid-like (and
arachnid-like) features which appear in certain of the
lower representatives of the vertebrate group, since
tendencies present in the ancestors which ultimately

148 THE AMERICAN NATURALIST [Vou. LIII

gave rise to both the Annelida and the Hemichordata
are quite likely to appear in both Annelida and Hemi-
chordata (or in forms descended from them, such as the
Arachnida and the Vertebrata).?

One of the chief difficulties in the way of reaching a
proper understanding of the mutual interrelationships
of the different lines of development is the attempt to
arrange these lines in the form of a dichotomously
branching tree drawn in one plane—which is almost as
impossible as the attempt to arrange all animals in a
single linear developmental series; for it’ must be borne in
mind that these different lines of development frequently
approach one another from different directions, so that
it would be necessary to represent their relationships by
a figure drawn in three dimensions, rather than in a
single plane. If this is done, it becomes easier to under-
stand that the line of development of the ‘‘Annelida,’’ .
for example, is paralleled (on different sides) by those
of several other groups, and that all of these lines of
descent may lead back to a common ancestry, or that
their points of origin may be near the point at which the
line of descent of the ‘‘ Annelida’’ arose.

In discussing the ultimate relationships of the Verte-
brata, Echinodermata, Mollusea, ete., the lack of inter-
mediate forms annectent between the different develop-
mental series, or connecting them with the supposedly
ancestral forms, has made the subject of their affinities
extremely speculative; and it is not until we come to the
consideration of arthropod phylogeny that the evidence
is at all satisfactory, and even here important gaps in
the developmental chain leave much to be desired.

As was previously stated, the members of the Myzosto-
mida-Onychophora group have developed many charac-
ters strongiy suggestive of arthropod affinities; but they

2 This statement should not be interpreted as implying that recent ver-
tebrates are descended from living Hemichordata, etc., but it is merely
intended to indicate that the Hemichordata have departed but little from

the probable ancestral condition of the Vertebrata, and the same holds true
for similar statements throughout this paper.

No. 625] THE EVOLUTION OF ARTHROPODS 149

have become too greatly modified along their own lines
of specialization in regard to those particular structures
most frequently used in comparative morphology to be
of much value for a phylogenetic study of the develop-
ment of the different parts of the body in the lower
arthropods. Among the ‘‘Annelida,’’ on the other hand,
we find some very promising material for such a study,
especially among the chetopodan annelids, such as the
Syllide (e. g., Dujardinia rotifera, ete.), which have seg-
mented appendages, while others of the group have de-
veloped structures no less interesting from the stand-
point of phylogeny, indicating that they have departed
but little from the ancestral condition of the arthropods.
The segmentation of the body of these annelids,* the
nature and relative positions of the heart and the di-
gestive, nervous and other systems, very readily lend
themselves to such an interpretation, and it is not a diff-
cult matter to derive the head region of a primitive
arthropod from that of the annelid type (Bernard, 1892),
or to derive the appendages of such an arthropod from
those of the annelidan type, as has been recently dis-
cussed by Borradale, 1917.

In connection with the discussion of the derivation of
the appendages of the lower arthropods from structures
comparable to the parapodia of the annelids, it may be
remarked that the attempt of Lankester, 1872, to derive
the Arthropoda more directly from the Rotifera, such
as the remarkable Pedalion mira (whose appendages
and the ‘‘arms’’ of the male Asplanchna he compares
to the movable spines of Triarthra and Polyarthra), has
not been productive of as important results as those ob-
tained from the comparison of the annelidan structures
with those of the arthropods. This, however, is merely
to be expected, since the annelids have developed far

3 Just as some annelids are many-segmented, while the bodies of others
are composed of fewer segments, it is quite reasonable to suppose that the
ancestors of .the arthropods exhibited a considerable range in the number
of segments composing their bodies—and even among the most primitive
arthropods there is a wide range in the number of segments composing
their bodies.

150 THE AMERICAN NATURALIST [Vor. LIII

more features in common with the lower arthropods than
have such primitive forms as the Rotifera. On the other
hand, the Annelida themselves (and hence ultimately the
Arthropoda also) are the descendants of Rotifera-like
(and Platyhelminthes-like) forebears, and it is quite
possible that certain rotifers might develop features
which later find opportunity for fuller expression in the
forms descended from them (e. g., the striated muscles
of Pedalion); but, since the general organization of a
rotifer’s body is not so similar to that of the lower
arthropods as is the case in the annelids in question, for
the present at least it seems preferable to regard the
slight resemblances between the appendages of the Rotif- .
era and the lower Arthropoda mentioned above as the
result of ‘‘convergence’’ (parallelism) in development
rather than to consider it as a precocious development of
structures later to be developed in the arthropod de-
scendants of ancestors ultimately arising from rotiferan
forebears. Iwould not utterly deny the latter possibility,
however, since it may be quite possible that arthropods
are to be derived more directly from Rotifera-like fore-
bears (e. g., Hexarthra polyptera, ete.) through forms
related to the Tardigrada and Nauplius-like ancestors;
but the great mass of evidence from comparative anat-
omy, embryology, etc., points to an ‘‘annelid ancestry”
for the Arthropoda, and until other hitherto undiscov-
ered forms have been found to indicate some other deri-
vation for the group, we are safe in assuming that the
‘*Annelida’’ represent as nearly as any known forms the
ancestral condition of the Arthropoda.

In taking up the consideration of the evolution of the
Arthropoda themselves, the question naturally arises as
to what arthropods have departed the least from the
probable ancestral condition of the group as a whole.
Some investigators would claim that since the Nauplius
larva is of such widespread occurrence among the lower
arthropods, that it represents an ancestral type; but it
must be borne in mind that a free-swimming larva is

No. 625] THE EVOLUTION OF ARTHROPODS 151

usually very greatly modified in adaptation to its own
mode of life and environmental conditions, and fre-
quently represents an interpolated stage having no great
-phylogenetic significance (in comparison with the devel-
opmental stages of the embryo). Furthermore, it is
extremely probable that the ancestral arthropods were
not of one single type at all, but doubtless differed very
greatly among themselves in size, the number of seg-
ments composing their bodies, ete., just as is the case
among certain annelidan groups, or as is the case among
the assemblage of lower arthropods comprising the most
primitive members of the group next to be discussed.
The assemblage of lower arthropods comprising the
Copepoda, Branchiopoda and their immediate relatives
may be referred toas the Copepoda-Branchiopoda group.
Its members include some of the most primitive of the
arthropods, and it may be regarded as representing as
nearly as any the forms giving rise to the different
arthropodan lines of development. The Ostracoda rep-
resent a line of development which branched off at an
early date, and should also be included in the group; but
they are not structurally so important as the Branchio-
poda, etc., for a phylogenetic study of the lines of descent
to which the ancestral arthropoda gave rise. The Cir-
repedia likewise represent a group which branched off
from this stem at an early date, but they are too degen-
erate, and have followed their own line of specialization
too far to be included among the primitive representa-
tives of the Copepoda-Branchiopoda group. The Trilo-
bita are very closely related to the Apodie and Branchio-
poda in general, for such trilobites as Nathorstia transi-
tans are somewhat annectent between the trilobites and
the branchiopod Opabina regalis described by Walcott,
1912, and such trilobites as Marella splendens are very
like certain Apodide, ete.; but the closest affinities of the
Trilobita appear to be with the group next to be consid-
ered, and although the trilobites have preserved many
very primitive features which might entitle them to a

152 THE AMERICAN NATURALIST [Vot LIH

position in the ancestral ‘‘Copepoda-Branchiopoda’’
group, it is preferable to consider them as members of
the Trilobita-Merostomata group, with which they have
much more in common.

The Trilobita-Merostomata group is composed of the
Trilobita, Eurypterida, and Xiphosura, with their imme-
diate relatives, and includes the forms which have de-
parted the least from the ancestral condition of the
arachnoids in general. The Trilobita are extremely
closely related to the Merostomata, some of which (such
for example as the fossil merostome Bunodes lunula,
which has been admirably restored by Patten, 1912, or
the cambrian merostomes, Sidneyia inexpectans, Emeral-
della brocki, ete., described by Walcott, 1911-1912) bear
well-developed antennæ very similar to those of the trilo-
bites; so that the division of the Arthropoda into ‘‘Te-
leiocerata’’ and ‘‘Chelicerata’’ by Heymons, 1901, or
into ‘‘Antennata’’ and ‘‘Chelicerota’’ by Boerner, 1909,
can not be strictly applied when we take these forms into
consideration. Walcott, 1912, considers that such mero-
stomes as Molaria spinifera are connected with the trilo-
bites through such intermediate forms as Nathorstia
transitans—a trilobite also related to the Branchiopoda.
Walcott also considers that the merostome Sidneyia rep-
resents a transition form between the trilobites and the
eurypterids, and that the merostomes Beltina and Sid-
neyia are related to the ancestors of living Xiphosura;
so that according to his views the trilobites are descended
from branchiopods, while the eurypterids are descended
from trilobites through such merostomes as his ‘‘Agla-
spina” and ‘‘Limulava,’’ from which living Xiphosura
are descended.

In discussing the lower arachnoid forms, it is neces-
sary to take into consideration the Pantopoda, which
have apparently retained certain features strongly sug-
gestive of crustacean affinities, while certain other fea-
tures suggest that they are related to the arachnoid
forms. Boerner, 1902, however, thinks that the Panto-

No. 625] THE EVOLUTION OF ARTHROPODS 153

poda are structurally quite far removed from the arach-
nids examined by him, and since he has made a very
extensive study of the different arachnoid forms, his
opinion should have considerable weight. Since their
line of development does not approach very closely to
those of the other forms here discussed, the study of the
Pantopoda is not of as great phylogenetic importance as
that of those forms which occupy a position annectent
between the other groups, or whose lines of descent ap-
proach those of the other groups. For the purpose of
the present paper, it is therefore sufficient to say that
the Pantopoda represent a highly aberrant group whose
line of descent branched off at an early date, somewhere
near the Trilobita-Merostomata group, and that they
have followed a widely divergent path of specialization.

The scorpions are descended from forms very like the
eurypterid members of the Trilobita-Merostomata group,
and such eurypterids as Glyptoscorpius occupy a posi-
tion annectent between the two groups. On the other
hand, the scorpions, together with the Pedipalpi, are in
many respects very like the ancestors of the higher
arachnids, so that they form an ancestral group, the
Scorpionida-Pedipalpi, intermediate between the Trilo-
bita-Merostomata group and the higher arachnids. In
the Scorpionida-Pedipalpi. group should be included the
closely allied pseudoscorpions and probably the Koene-
nia-like forms and the Solifuge also.

Of the higher groups of arachnids, the spiders (Ara-
nee) are apparently quite closely related to the ambly-
pigid (tarantulid) branch of the Pedipalpi, while the
Phalangidea (Opiliones) and Acarina are more closely
related to the pseudoscorpions and Solifugæ, although it
has been suggested that the Cryptostemmatide occupy a
position intermediate between the Pedipalpi (to which
they are somewhat more closely allied) and the Phalangi-
dea. The Linguatulida have been placed near the mites
by many recent authorities on account of the supposed
resemblance of their larve to such short-bodied mites

154 THE AMERICAN NATURALIST [Vou. LIII

as Phytoptus, ete., and the apparent resemblance of the
adults to eriophyid mites, so that provisionally, at least,
the Linguatulida may be regarded as strongly aberrant
mites, while the Tardigrada are probably not related to
them, but to the Myzostomida, as has been previously
discussed.

Since the arachnoid path of evolution has led off in a
direction widely divergent from the path followed in the
development of the Insecta, it is very difficult to under-
stand how Thorell came to the conclusion that such
highly developed arachnids (i. e., those occupying a posi-
tion far along the divergent line of development) as the
Solifuge are intimately related to insects. Further-
more, since the trilobite trend of development leads off
toward the merostomes and the divergent evolutionary
path of the arachnoid forms, it is necessary to search
further back than the trilobites for a group standing
more nearly in the direct line of development eventually
resulting in the evolution of the insectan type, and for
this purpose the study of the branchiopod representa-
tives of the Copepoda-Branchiopoda group is much more
valuable.

The members of the Copepoda-Branchiopoda group
which seem to be the nearest to the stem forms at the
base of the line of descent which ultimately leads up to
the insect type of development are the Notostraca (Apo-
did) and anostracan Branchiopoda, which are likewise
very closely related to the trilobites, so that certain an-
cestral features are to be found in the trilobites also,
having been inherited from their common forebears; but,
as was stated above, the trend of trilobite development
is toward the production of the eurypterid and arachnoid
type of development, and therefore leads away from the
line of development which eventually results in the pro-
duction of the insect type. Walcott, 1912, agrees with
Bernard, 1892, in regarding the Apodide as among the
lowest representatives of the Arthropoda (although cer-
tain copepods are also extremely primitive) and suggests

No. 625] THE EVOLUTION OF ARTHROPODS 155

that the fossil annelids, Canadia spinosa (in which the
` head is bent down ‘‘so that the mouth faces posteriorly”?
in the position assumed by Bernard, 1892, to be that of
the annelids which gradually took on the character of
head region leading up to the arthropod type), and the
Crustacea ‘‘were derived from the same general type of
animal.” The Copepoda represent a line of develop-
ment which branched off near that of the Branchiopoda,
at the base of the arthropod stem; and the Argulide
(which are grouped with the Copepoda by Calman, 1909)
are regarded by some authorities as annectent between
the Copepoda and the Branchiopoda. The Ostracoda
are related to both the conchostracan and cladoceran
Branchiopoda (following Calman’s classification) and
the ancestors of the ostracods doubtless arose from forms
intermediate between the Cladocera and Conchostraca.
The Cirrepedia are apparently descended from ancestors
related to both the Ostracoda and Copepoda, and their
line of development branched off at an early date to fol-
low their own strongly aberrant part of development.

‘Such anostracan branchiopods as the fossil Opabina
regalis, whose structure according to Walcott, 1912, ‘‘is
very suggestive of an annelidan ancestor,’ and such
notostracan branchiopods as the fossil Burgessia bella
(which has sessile eyes and hepatic glands in a‘ carapace
resembling that of Lepidurus) serve to indicate what the
first arthropods were probably like, and they occupy a
- position near the base of the stem-forms whose lines of
development were eventually to produce the insectan
type of arthropod. The fossil notostracan branchiopod
Waptia occupies a position annectent between the above-
mentioned branchiopods and the malacostracan group
next to be considered.

The leptostracan (phyllocarid) group occupies a po-
sition intermediate between the rest of the Malacostraca
and the branchiopods described above. They have also
carried over from their common branchiopod ancestry
certain features likewise inherited by the trilobites; but.

156 THE AMERICAN NATURALIST (Vou. LII

as was previously stated, the trilobites do not stand in
the direct line of descent of the Leptostraca, and those
characters which they possess in common were inherited
from their common branchiopod ancestry, and can not
be interpreted as indicating that the trilobites represent
the ancestral forms giving rise to the lepostracan type.
The fossil leptostracan Hymenocaris is evidently related
to the fossil branchiopod Waptia (which occupies a posi-
tion intermediate between the branchiopods and Lepto-
traca), but Hymenocaris is clearly a leptostracan, and
resembles such living forms as Nebalia, while the fossil
leptostracans Carnavonia and Tuzoia resemble such liv-
ing Leptostraca as Nebaliopsis typica in the character
of the carapace, etc. The closer affinities of the fossil
Ceratiocaride, ete., have not been determined, due to the
imperfect preservation of the limbs, ete., but they clearly
belong to the leptostracan group. There is much to be
said in favor of including the Leptostraca in the next
group of the Malacostraca to be considered, but from the
standpoint of a phylogenetic study it is preferable to
consider the Leptostraca (together with other primitive
forms not yet described) as nearer the ancestral forms
from which the other Malacostraca were derived.

A further stage of development is represented by the
Anomostraca-Cumacea group which includes the Synca-
rida and a portion of the Peracarida of Calman’s classi-
fication, together with their immediate relatives. The
Anomostraca (Anaspidacea and Bathynellacea of Chap-
pius, 1915), Mysidacea and Cumacea are very closely
interrelated, and all of them exhibit affinities with the
Leptostraca, so that the members of the leptostracan
group might well be included here also; but they have
been treated as a separate group, to emphasize the fact
that they occupy a position annectent between the
Branchiopoda and the Malacostraca (with which their
strongest affinities lie). Although the members of Ano-
mostraca-Cumacea group are extremely closely related
to the Tanaidacea, the closest affinities of the Tanaidacea

No. 625] THE EVOLUTION OF ARTHROPODS 157

are with the Isopoda (and Amphipoda), so that it is pref-
erable to consider them with the latter group. The Ano-
mostraca-Cumacea group is of the greatest phylogenetic
importance, since its members have departed as little as
any known forms from the probable ancestral condition
of the higher Crustacea, Insecta and ‘‘Myriopoda’’
(sensu lato).

The Mysidacea have retained some primitive charac-
ters indicating their connection with the Leptostraca-
- like forms which preceded them, and they are quite like
the ancestors of the eucaridan (euphausiacean and deca-
podan) members of the higher crustacean groups. They
are also probably related more remotely to the ancestors
of the aberrant hoplocaridan (stomatopodan) line of de-
velopment, and through such forms as the Cumacea they
are connected with the ancestors of the Tanaidacea (and
therefore of the Isopoda also). They are not so impor-
tant for a phylogenetic study of the insects, etc., how-
ever, as the Anomostraca and Cumacea (with the Tanai-
dacea) are. The Cumacea occupy a position interme-
diate between the Mysidacea and the Tanaidacea, being
somewhat more closely allied to the latter. They are
also related to the Anomostraca (Syncarida), as is true
of the Mysidacea, the interrelations of the different
members of the group being rather complicated.

From the standpoint of the study of the phylogeny of
the insects and their relatives the Anomostraca and Cu-
macea (together with the Tanaidacea) are by far the
most important forms, since the ancestors of the insects
and their relatives were doubtless descended from forms
closely related to the Anomostraca, Cumacea, and Tanai-
dacea. Of these three, the Anomostraca are apparently
the most ancient (fossil remains of the others have not
yet been found), and have doubtless departed as little
as any from the ancestral forms which were eventually
to give rise to the isopods, insects and ‘‘myriopods.”’
The fossil Pleuroearide (e. g., Acanthotelson, ete.) are
nearer to the living genera Koonunga, Anaspides, Para-

158 THE AMERICAN NATURALIST [Von LIH

naspides, etc., while the fossil ‘‘Gampsonychide’’ (e. g.,
“Gampsonyx,” Paleocaris and Gasocaris) are nearer
the living genus Bathynella. Such fossil forms as Prea-
naspides found in the Carboniferous rocks is extremely
like the living Anaspides which has apparently pre-
served many ancestral characters, but little modified, to
the present time. The Anomostraca are related to the
Leptostraca, but no forms intermediate between them
and the Leptostraca have yet been described, and it is
possible that the line of descent of the Anomostraca leads .
back to the branchiopods through Leptostraca-like forms
not yet discovered. Superficially, at least, such slender
branchiopods as Yohoia tenuis, ete., resemble certain
members of the Anomostraca, and it is possible that the
slenderer, more cylindrical Anomostraca, such as Bathy-
nella, may have inherited the tendency toward the slen-
der form of body from anostracan branchiopods of the
Yohoia type. In Bathynella the eyes have become com-
pletely lost, but in Koonunga sessile eyes are found and
their presence suggests that sessile-eyed forms may have
developed from the Koonunga type. In Anaspides the
eyes are stalked.

From ancestors occupying a position intermediate be-
tween the Anomostraca and Cumacea (and also related
to the Mysidacea) have arisen the lines of descent leading
to the isopod Crustacea, Insecta, and ‘‘Myriopoda’”’ (s.
1.). The Tanaidacea (Chelifera) which occupy a posi-
tion near the base of the isopod stem are very closely
related to the Anomostraca, Cumacea and Mysidacea,
and, together with the Isopoda and Amphipoda (which
are descended from ancestors very similar to them), they
might be included in the Anomostraca-Cumacea group;
but if the Isopoda-Amphipoda group is considered sepa-
rately, the Tanaidacea must be included in the latter
group, since their closest affinities are with the Isopoda.
The Amphipoda are quite closely related to the Isopoda,
and their ancestors may also have arisen from forms
intermediate between the Anomostraca and Cumacea

No. 625] THE EVOLUTION OF ARTHROPODS 159

(and also related to the Mysidacea), so that the sessile-
eyed character occurring in the group might be regarded
as a retention of the tendency toward the formation of
sessile eyes exhibited by such primitive forms as Koo-
nunga, while the slender body form present in such Am-
phipoda as the caprellids, Rhabdosoma, ete., may possi-
bly be due to the retention of the tendency toward the
slender form of body (such as that present in the more
primitive Bathynella) in forms which are otherwise rather
highly modified. The Isopoda-Amphipoda group origi-
nated very close to the point of origin of the insect line
of development, and the two lines have paralleled one
another extremely closely. Since the’ members of the
TIsopoda-Amphipoda group have not travelled so far
along the path of specialization in following the same
developmental road with the insects, they have retained
many primitive features characteristic of the ancestors
of the insects (and ‘‘myriopods’’), and such forms as
Apseudes are particularly interesting for a phylogenetic
study of insects and their immediate relatives.

The Symphyla-Pauropoda group (composed of such
forms as Scolopendrella, Scutigerella, Pauropus, Eury-
pauropus, and their immediate relatives) contains the
forms which appear to be very near the base of the
‘‘myriopod’’ stem, and which have retained a great num-
ber of features characteristic of the ancestors of insects,
so that a study of the structures of the Isopoda-Amphi-
poda group and the Symphyla-Pauropoda group are of
the greatest importance for a proper conception of the
nature of the first insects to be evolved. The Symphyla-
Pauropoda group probably also arose from forms occu-
pying a position intermediate between the Anomostraca
and Cumacea, and likewise closely related to the Tanai-
dacea which originated from similar forebears. Such
Anomostraca as Bathynella have not departed far from
the ancestral condition of the Symphyla-Pauropoda
group, and although they have developed many modifi-
cations along their own line of specialization, they are as

160 THE AMERICAN NATURALIST [Vou. LIII

near as any known forms to the ancestors of the Sym-
phyla, ete. The Symphyla-Pauropoda group in turn has
departed but slightly from the ancestral condition of the
‘‘Myriopoda’’ as a whole, although the ancestral ‘‘My-
riopoda’’ comprised forms with bodies composed of
more numerous segments as well as those made up of
fewer segments. From ancestors similar to the mem-
bers of the Symphyla-Pauropoda group one line of de-
velopment has led to the chilopod type of myriopod,
while the other has led to the diplopod type. From their
ancestors related to the members of the Symphyla-Pau-
ropoda group, the Chilopoda have carried over many
characters also inherited by the ancestors of insects, so
that a structural study of the Chilopoda is of consider-
able value from the standpoint of insect phylogeny (as
is true to a lesser degree of the Diplopoda also).

As was stated above, the ancestors of the Insecta were
related to the members of-both the Isopoda-Amphipoda
group (including the Tanaidacea) and the Symphyla-
Pauropoda group, so that the lines of descent of all three
groups (insects, isopods and Symphyla) doubtless had
a common origin in forms intermediate between the Cu-
macea and Anomostraca (and also related to the Mysi-
dacea), and all of the three groups have inherited from
their common ancestry many characters also carried
over in the lines of development of the other two of the
three groups in question. The common ancestors of the
three groups just mentioned (insects, isopods and Sym-
phyla) were not of any one single type, but doubtless
differed quite markedly among themselves in the number
of segments composing their bodies, the slender or
stouter and flatter character of the body and other fea-
tures. Some of them were more like the Tanaidacea,
while others were more like Bathynella and other mem-
bers of the Anomostraca, ete., and this should be clearly
borne in mind in attempting to determine what the an-
cestors of the insects, ete., were like; for the greatest
obstacle to arriving at the realization of the true nature

No. 625] THE EVOLUTION OF ARTHROPODS 161

of the ancestors of insects and their relatives has been
the attempt to derive them all from one type of creature
—which is manifestly impossible, since even the lowest
representatives of any group differ markedly among
themselves, and their ancestors also must have differed
markedly among themselves (although not to such a
great extent as their progeny do).

Although such Anomostraca as Bathynella have be-
come specialized along their own lines of development,
they have retained many features which suggest what
some of the ancestors of the insects and Symphyla must
have been like, and I think it very probable that the an-
eestors of Scolopendrella and the Protura were quite
similar in many respects to Bathynella, while other ap-
terygotan insects, such as Machilis, have carried over
more characters from the tanaidacean side of their com-
mon ancestry. Therefore, if we accept the idea that
some of the common ancestors of insects, isopods and
Symphyla occupied a position intermediate between the
lines of development of the Anomostraca and the Cu-
macea-Tanaidacea, and differed a little less among them-
selves than the Anomostraca do from the Cumacea-
Tanaidacea, it becomes perfectly clear that some aptery-
gotan insects could inherit from the tanaidacean side of
their common ancestry characters which also appear in
the isopods which are derived from Tanaidacea-like fore-
bears; while on the other hand, other apterygotan insects
could inherit from the Bathynella side of their common
ancestry certain characters which also appear in the
Symphyla or other forms descended from Bathynella-
like forebears.

The Protura (such as Acerentomon, Eosentomon, ete.)
are the most primitive representatives of the Insecta,
and have inherited from their common ancestry many
features also preserved in the ‘‘Myriopoda’’; and the
embryological development of the apterygotan group to
which they belong has much in common with that of the
**Myriopoda,”’ as has been pointed out by Philiptschenko,

162 THE AMERICAN NATURALIST (Vou. LII

1912, Lignau, 1911, Chamberlain, 1917, Heymons, and
others. The retention of the stumps of three pairs of
legs on the abdominal region (in addition to the three
pairs of thoracic legs) at first caused some zoologists to
doubt that the Protura are really insects (since the idea
that such forms with vestigial abdominal legs could not
be true ‘‘hexapods’’ if they had more than six limbs
seemed to stand in the way of their realizing the true
insectan nature of the Protura), but the overwhelming
evidence of their structural organization has convinced
all recent entomologists that the Protura are true insects.
As pointed out in a recent paper (Crampton, 1916) the
Protura are quite closely related to such other Aptery-
. gota as Tomocerus; and, with the Entomobryids and
Sminthurids, they constitute the non-styli-bearing divi-
sion of the Apterygota.

Of the styli-bearing Apterygota, the next group to be
considered, which may be referred to as the Campodeoid
group, comprises the Rhabdura (e. g., Campodea), the
Dicellura (e. g., Projapyx, Japyx, ete.) and their imme-
diate relatives. Dicellura, such as Projapyx, Anajapy2,
etc., have segmented cerci, and occupy a position inter-
mediate between the Rhabdura, such as Campodea, and
the other Dicellura, although their closest affinities are
clearly with the Dicellura. The Campodeoid group,
whose members have entognathous mouth parts and ves-
tigial abdominal legs suggestive of the proturan struc-
tures, occupy a position intermediate between the lower
apterygotan Protura and the higher apterygotan forms,
such as Nicoletia, Lepisma, ete., which also belong to the |
styli-bearing apterygotan subdivision which includes the
Campodeoid group as well (Crampton, 1916). The Cam-
podeoid group, while inheriting certain features from the
symphylan side of their common ancestry, have inherited
in addition certain other features more typical of the
crustacean side—which likewise reappear in the isopod-
amphipod descendants of their common ancestors.

The Lepismoid group, composed of the lepismids,

No. 625] THE EVOLUTION OF ARTHROPODS 163

machilids, and their immediate relatives, is quite closely
connected with the Campodeoid group in the styli-bear-
ing subdivision of the Apterygota; but their mouth parts
are ectognathous, and in their general organization they
approach remarkably closely to the lower Pterygota;
so that they may be said to occupy a position annectent
between the lower Pterygota and the Campodeoid group.
The members of the Lepismoid group seem to have in-
herited more characters from the crustacean side of their
common ancestry than from the symphylan side, while
the members of the Proturan group seem to have inher-
ited more characters from the symphylan side, and the
members of the Campodeoid group appear to partake to
some extent of characters occurring in both the crusta-
cean (isopod) and symphylan sides of their common
ancestry.

It might be possible to explain the presence of both
crustacean (isopod) and symphylan characters in the in-
sectan stem by supposing that the crustacean, insectan
and symphylan ‘‘currents’’ in the ‘‘onward flow of life,’’
although acquiring more and more of a distinct indi-
viduality as their ‘‘waters’’ emerge from the common
stream at their source, nevertheless have an intermin-
gling or commingling of contiguous waters as they flow
side by side, before ultimately diverging too greatly for
such an intermingling. This idea, however, might in a
sense be interpreted as meaning that the Symphyla-like
insects were descended from Symphyla, and the Crus-
tacea-like insects from Crustacea (i. e., isopod Crus-
tacea), whereas insects as a whole were probably not
‘“polyphyletic,’’ but all insects were derived from a
common ancestral source. The forms composing this
common ancestral source, however, differed among them-
selves very greatly, although the amount of divergence
was probably not too great to prevent their being grouped
in a single class—or possibly even in a single subclass
or order. In this ancestral-insectan group, there were
doubtless isopod-like insects which resembled the most

164 THE AMERICAN NATURALIST [Von. LIII

insect-like representatives of the ancestral isopods, while
the Symphyla-like members of the ancestral-insectan
group must have resembled the most insect-like repre-
sentatives of the ancestral Symphyla. In other words,
at the common level at which the lines of descent of the
isopods, insects and Symphyla originated, some of the
ancestral insects (which differed greatly among them-
selves) occupying the ‘‘hereditary territory’’ contiguous
to that of the ancestral Symphyla would inherit certain
developmental tendencies in common with or similar to
those also inherited by certain Symphyla; and similarly,
some of the ancestral insects occupying the ‘hereditary
territory” contiguous to that of the ancestral isopods
would inherit certain developmental tendencies similar
to those of certain isopods and the same principle would
apply to successively larger, as well as to the smaller
groups in any evolutionary study. According to this
view, certain developmental or ‘‘inherent’’ tendencies
exhibited by the isopods or myriopods might also appear
in insects if the opportunity of manifesting themselves
should arise, and this would merely imply that these
tendencies were inherited from an ultimately common
ancestry, rather than that some insects were descended
from isopods while other insects were descended from
Symphyla, ete. Some evolutionists might object to the
use of such terms as ‘‘inherent tendencies’’ on the ground
that they savor too strongly of ‘‘vitalism’’; but, so far
as I can see, the expression ‘‘inherent tendencies’’ means
much the same thing as a part of ‘‘heredity,’’ and one
implies no more of a predilection toward vitalism than
the other does.

Although their closest affinities are with the Cam-
podeoid group and the Apterygota in general, certain
members of the Lepismoid group are structurally remark-
ably similar in many respects to such primitive Ptery-
gota as the stone-flies and may-flies, so that Handlirsch,
1906, who has completely disregarded the close interrela-
tionships of the Apterygota, and their evident ancestral
character (with reference to the winged insects) in his

No. 625] THE EVOLUTION OF ARTHROPODS 165

attempt to derive the Pterygota more directly from trilo-
bites, is forced to assume that the lepismids may represent
degenerate Pterygota! Their whole sturctural organiza-
tion clearly proclaims in no uncertain terms that the
closest affinities of the lepismids are with the rest of the
Apterygota, with which they are connected by inter-
mediate forms, and a careful study of the comparative
anatomy and embryology of the Apterygota, ‘‘ Myrio-
poda” and Crustacea can result in no other conclusion
than that the Apterygota have departed as little as any
known forms from the condition characteristic of the an-
cestors of the Pterygota. The lepismids are therefore no
more to be considered as degenerate Pterygota, than apes
are to be considered as degenerate men—unless one re-
verses the whole scheme of evolution; and under such
conditions there would be nothing to prevent any one from
assuming that trilobites are degenerate lepismids, or any
other equally improbable reversing of the evolutionary
sequences !

In connection with the supposedly ‘‘degenerate’’ con-
dition of the Apterygota, I would take issue with the im-
plication carried in such statements as that by Tothill,
1916 (p. 376), who would claim that the Apterygota ‘‘are
highly specialized animals as indicated by the frequent
reduction of mouth parts, visual organs, tracheæ, ete. ;
and by the development of peculiar structures such as the
caudal spring and collophore.’’ In the first place, it is
inadmissible to judge the ancestral character of any
group by the condition of its most highly specialized
members, as Tothill appears to do in the case of the
Apterygota, since any arthropodan group, no matter how
low it may be in the seale of development (e. g., Cope-
poda, ete.) may include certain members which have be-
come very highly specialized along their own lines of de-
velopment without affecting the general position of the
group as a whole; and in a phylogenetic study we must
consider the most primitive representatives of the group,
rather than the most highly specialized ones, if such a
study is to yield any tangible results. If Tothill had

166 THE AMERICAN NATURALIST [VoL LIH

therefore considered such lowly organized Apterygota as
- Eosentomon, Anajapyz, ete., instead of the highly special-
ized Anurida, Sminthurus, ete., I am sure that his opinion
of the “‘degenerate’’ condition of the Apterygota as com-
pared with the Pterygota would have been quite the
opposite of that expressed in his paper. Furthermore,
there are practically no arthropods known which are
primitive in all respects, and, as is the case throughout
the whole realm of zoology, forms which have retained
many features in an exceedingly primitive condition may
be very highly specialized in other respects; so that one
must take into consideration the composite primitive
features of the group as a whole; and, just as the most
primitive members of the Pterygota are studied in an
attempt to determine their ancestry, so the most primi-
tive members of the Apterygota must be considered in
such a phylogenetic study.

Even in the matter of the nature of their eyes, such
forms as Machilis (which are related to Lepisma) can
hardly be called ‘‘degenerate,’’ and in the face of the
fact that in the trilobites themselves there occur at least
three types of eyes—‘‘isolated eyes or ocelli, aggregate
eyes of biconvex lenses, and compound eyes”? (Tothill,
p. 321, quoted from Lindstrom, 1901), it is very improb-
able that the type of eyes found in Lepisma are of a
higher type than the compound eyes of the Pterygota.
As far as their mouth parts are concerned, I find the
lepismids much more primitive than the Pterygota (with
the possible exception of nymphal ephemerids) and Boer-
ner, 1908-1909, has called attention to crustacean struc-
tures so similar to those found in the maxille, etc., of
apterygotan insects, that there can be no doubt that the
mouth parts of the Apterygota in general instead of
being ‘‘degenerate’’ have retained many more primitive
features than those of most lower Pterygota.

As far as the number of abdominal segments is con-
cerned, some Apterygota, instead of having fewer seg-
ments, have even retained twelve, and in these forms,
such as the Protura, there is also a postembryonie in-

No. 625] THE EVOLUTION OF ARTHROPODS 167

crease in the number of segments (from nine to twelve
in the abdomen) comparable to the increase of segments
in the ‘‘Myriopoda,’’ so that Tothill’s statement that
‘<in the Hexapoda numerous investigations have shown
the segments arise only during the egg stage’’ does not
hold in the case of the Protura. There is also one other
point in Tothill’s paper which might easily lead to error
unless properly explained: namely, that his discussion of
the nature of the appendages of the abdomen in a
‘‘larval’’ Stenodictya is based upon a figure taken from
Handlirsch’s book, the supposition being that it repre-
sents the restoration of an actual fossil larva, whereas in
reality the figure is purely a figment of Handlirsch’s
imagination, for no known insects have biramous ab-
dominal legs, and even the supposedly biramous condition
of such specialized structures as the maxille of insects is
now thought to be a secondarily acquired feature, and not
a retention of an originally biramous condition (Borra-
dale, 1917). Tothill’s suggestion of a derivation of
winged insects directly from Chilopoda (which represent
a side branch from the symphyloid main stem of myriopod
development) without reference to the apterygotan forms
is open to all of the objections raised against deriving
winged insects from apterygotan forms without having
any of the advantages of the latter hypothesis, and if
_ the latter is untenable, the idea of deriving winged insects
from chilopods is infinitely more so!

` Despite the fact that trilobites are on a divergont
branch leading away from the main line of insectan de-
velopment (i. e., leading off to the arachnoid develop-
ment) Handlirsch, 1906, would derive winged insects
directly from trilobites, wholly ignoring the Apterygota,
Symphyla, Tanaidacea, and all of the other anatomically
intermediate forms—which would be exactly on a par with
an attempt to derive the ‘‘Nordic’’ race of men directly
from lemurs (or rather from cats, whose line of develop-
ment has deviated from the main line of evolution lead-
ing to the development of the human type) wholly ignor-
ing the Mongolians, Australoids, Neanderthaloids,

168 THE AMERICAN NATURALIST [Vou. LIII

Heidelberg man, Pithecanthropus, the great apes, and
all.other anatomically intermediate types! His line of
argument is somewhat as follows: winged insects oc-
curred at an extremely early period, and no fossil Aptery-
gota dating back to so ancient a period has yet been dis-
covered; therefore Apterygota are more probably a re-
cent degenerate offshoot, rather than forms standing
more nearly in the line of development of winged insects—
a line of reasoning which caused the earlier Coleopter-
ologists to reverse the evolutionary sequence and attempt
to derive true beetles from the snout beetles, until further
discoveries brought to light the fact that true beetles were
geologically as ancient, or more ancient, than the snout-
beetle type, which comparative anatomy clearly showed
must have been derived from, and therefore could not be
ancestral to, the true beetles! As experience has shown,
the paleontological evidence, which at best is of a most
fragmentary and incomplete nature, must supplement
that of comparative anatomy (of adults or embryos) —
and even in the case of the paleontological evidence it de-
pends wholly upon comparative anatomy here also; and
furthermore many fossils were themselves as highly
specialized along their own lines of development as the
most primitive living forms are (some of which have re-
tained just as many ancestral characters and are as little
modified in certain respects as those forms which fell by
the wayside at an early date). Paucity in numbers of
individuals among the Apterygota, their usually smal!
size and fragile nature, have all contributed to make their
fossil remains extremely rare, and under these conditions
the lack of remains from earlier strata can not offset the
weighty argument of comparative anatomy and embryol-
ogy in favor of regarding them as the nearest representa-
tives of the type ancestral to winged insects.

As for deriving winged insects directly from trilobites
on the ground of the faint resemblance of trilobites to
insects in regard to their possession of a certain type of
eye structure, antenne, and lateral projections of the
tergal region (woefully inadequate resemblances in com-

No. 625] THE EVOLUTION OF ARTHROPODS 169

parison with the multitude of resemblances between in-
sects and their real ancestral forms), it may be said that
these same structures are likewise shared by such fossil
merostomes as Bunodes lunula and on precisely the same
grounds, insects should be derived from merostomes also
(a manifest impossibility) since these have the same
ancestral qualifications of great antiquity, and they
possess the trilobite type of antenne, eyes and lateral
tergal projections! When one studies the embryological
development of insects, however, it is evident that their
ancestors had two pairs of antenne instead of the one
pair apparent in trilobites, and the insectan type of head
is nothing like that of a trilobite in which the head region
is not set off by a marked constriction with well-defined
mandibles, maxille and underlip of the insectan type,
while the head region and mouth parts of isopod and
amphipod Crustacea, ete. (with their two pair of an-
tenn, insectan type of head, mandibles, maxille and
underlip), are clearly similar in character to what the
ancestors of insects must have been like, and the same
holds true of the legs and terminal appendages, ete., in
these Crustacea. Therefore, as far as comparative
anatomy is concerned the Crustacea, with their progeny
the Symphyla, ete., are, beyond any possibility of doubt,
the nearest forms to the ancestors of insects in general,
and this is also borne out by embryology, which, how-
ever, can not be applied in the case of the trilobites; so
that here we must depend largely upon comparative
anatomy, whose verdict is unmistakably in favor of the
Crustacea, Symphyla and Apterygota as the ancestral
forms leading up to the pterygotan type, and is unmistak-
ably against considering the trilobites anywhere near the
immediate ancestors of winged insects or even in their
direct line of descent. On this account, it is most
astonishing that nearly all recent writers (Schuchert, 1915,
Ruedemann, 1916, Lull, 1917, ete.) have accepted with-
out reservation such startlingly revolutionary ideas as
those proposed by Handlirsch—and upon such meagerly
insufficient grounds when one looks into the subject at

170 THE AMERICAN NATURALIST [Vor. LIII

all! Such implicit faith in this age of skepticism speaks
volumes for the weight of Handlirsch’s authority among
paleontologists, but the true morphologist prefers the
direct evidence of his own observation to any ‘‘petitio ad
auctoritatem” especially when such startlingly revolu-
tionary ideas as those which Handlirsch proposes are
based upon no firmer foundation than a vague resem-
blance which will not even bear the test of close scrutiny.

When one turns to the published figures of the earliest
fossil insects for some light upon the nature of their body
structures, his eye is met by a dreary succession of disem-

bodied wings, and in the rare instances in which the body
parts are also figured, only the vague outlines are given
with a nonchalant disregard for the vital details so neces-
sary for any phylogenetic study; and one can not help
wondering what impression the ‘‘pterophilous’’ paleon-
tologists would have of their subject if the tables had
been reversed and they had been presented with merely
the vaguest outlines of a series of wings containing no
veins or other important structures, in the expectation
that such figures would be of any value for a phylogenetic
study! Furthermore, many living ‘‘synthetic’’ types
are quite devoid of wings (as it true of immature forms
also) and the study of these forms is in some cases even
more important than that of the wing-bearing ones (e. g.,
Timema, Grylloblatta, nymphal Plecoptera, Lepisma,
etc.), but how are we to compare them with a series of
disembodied wings? So far as one can judge from the
figures of fossil insects, we have living to-day certain
lowly organized forms which are in many respects just as
primitive as these fossil forms (which are also specialized
to some extent) and when the paleontologist returns
again and again to a comparison with living forms for
an interpretation of fossil structures, the suspicion be-
comes a conviction that a study of the primitive char-
acters of various lowly organized living insects is just as
instructive from a phylogenetic point of view, and is
infinitely more satisfactory than a laborious reconstruc-
tion of fossil fragments.

No. 625] THE EVOLUTION OF ARTHROPODS FIL

The different theories concerning the origin of the
wings of pterygotan insects were discussed in a recent
paper (Crampton, 1916) in which it was pointed out that
it is possible to consider that the wings of insects were
derived from paranotal outgrowths of the tergal region of
apterygotan forms, Crustacea, ete., which are ultimately
homologous with the paranotal outgrowths of the trilo-
bites, without attempting to derive the wings from these
trilobitan structures without the intermediation of other
ancestral forms. Not only do the lepismids exhibit para-
notal structures (lateral tergal outgrowths) which are
homologous with the precursors of wings, but the lepis-
moid forms (Lepisma, Nicoletia, Machilis, ete.) approach
remarkably closely to the pterygotan type in many re-
spects, and may be considered as annectent between the
remainder of the Apterygota and the lower Pterygota.

The lowest representatives of the Pterygota, or winged
insects, constitute the Perlid-Ephemerid group, composed
of the Plecoptera, Ephemerida, and their immediate rela-
tives. The modern representatives of the group are in
many respects fully as primitive as certain of their fossil
relatives, although it is necessary to turn to some such
extinct forms as the ‘‘Protephemeroidea’’ and Palæo-
dictyoptera to find the connecting forms annectent be-
tween the Plecoptera and the ephemerids. The imma-
ture Plecoptera are remarkably Similar to lepismids in the
nature of the head outline, mouth parts, thoracic sclerites,
ete. (Crampton, 1917a), and even in regard to their
terminal abdominal structures the lepismids are very like
Plecoptera (Crampton, 1918a), but the Plecoptera have
lost the median terminal filament, which, however, is still
retained in the ephemerid members of the group. The
ephemerids, and the Odonata, represent somewhat aber-
rant types of development which branched off at an early
date to follow their own paths of specialization, although
they have not proceeded very far along this road. The
Plecoptera, on the other hand, have carried over in their
line of inheritance a great many characters which were
to become further developed in the higher groups of in-

172 THE AMERICAN NATURALIST [Von LIII

sects, and they appear to have departed as little as any
from the ancestral condition of these groups, so that they
are as important as any synthetic types, with the possible
exception of the Paleodictyoptera, for a phylogenetic of
winged insects in general. The great antiquity of fossil
Plecoptera is also in harmony with the idea that the
Plecoptera are quite like the ancestors of the higher
forms, and since the anatomical and phylogenetic data
are in complete harmony in this respect, we are justified
in assuming that the Plecoptera have departed as little
as any forms from the ancestral condition of the groups
next to be considered.

The Plecoptera, embiids, and Dermaptera originated
from essentially similar ancestors, which were not far
removed from present-day Plecoptera, and their lines of
descent have followed a common developmental road for
a considerable distance, before first the embiids, and a
little later the Dermaptera branched off to follow their
own paths of specialization (Crampton, 1917a). The
Hemimerus-like forms branched off from the Dermapteron
stock at an early date, and a little later, the Coleopteron
type was differentiated. The Strepsiptera were possibly
differentiated from a similar stock still later. The terms
‘‘earlier’’ or ‘‘later’’ as used above are employed in the
sense of indicating the relatively lower or higher level
along a line of development, at which a group branched
off, and is based upon the comparative anatomical primi-
tiveness of the group under consideration. In the case
of the Coleoptera, Handlirsch maintains that they are
paleontologically older than the Dermaptera, and if sub-
sequent findings should corroborate this view, it would
be necessary to search for the origin of the Coleopteron
line of development lower down on the Plecopteron stem
than the point at which the Dermaptera branched off to
follow their own path of specialization, but the Dermaptera
are so much more lowly organized than the Coleoptera,
to which they are anatomically very similar (see also
Crampton, 1918b), that I am inclined to believe that the
lack of earlier Dermapteron remains is due to the incom-

No. 625] THE EVOLUTION OF ARTHROPODS 173

pleteness of the fossil record, rather than to the absence
of Dermapteron forms antedating the Coleoptera.

The Isoptera, blattids and mantids seem to have
originated from a stock similar to the members of the
Plecopteron group mentioned above (Crampton, 19174)
and they apparently branched off at a very early date to
follow their own developmental road for a short distance
before each of the three separated to follow its own path
of development. The Isoptera are anatomically inter-
mediate between the members of the Plecopteron group.
and the rest of the blattid group (with which the Isoptera
seem to have somewhat stronger affinities than with the
members of the Plecopteron group, although they are
related to the embiids and Dermaptera quite closely).
This might be taken to indicate that the Isoptera are
more primitive than the blattids, as is borne out by cer-
tain of their anatomical features; but on the whole, the
blattids seem to be somewhat more lowly organized and,
according to Handlirsch, the Isoptera are paleonto-
logically much younger than the blattids. It is quite
probable that the Zoraptera described by Silvestri, 1913,
are an offshoot of the isopteron stock.

The orthopteroid insects, grylloblattids and phasmids
were descended from ancestors very similar to Gryllo-
blatta recently described by Walker, 1914, and such
phasmids as Timema are also very near the base of the
orthopteroid stem. These insects have inherited many
characters from the plecopteroid side of their ancestry,
and they also share many features in common with the
blattoid group mentioned above (see Crampton, 1917a).
Their line of descent is apparently ultimately traceable
to a plecopteroid ancestry (as is probably also the case
with the blattoid forms), but their line of development
branched off very near that of the blattoid group, and
they continued to parallel the path of development of the
latter group for a considerable distance before diverging
along their own branch of specialization. The gryllo-
blattids seem to be somewhat closer to the ancestors of
the gryllids and ‘‘locustids,’’ while the phasmids may be

174 THE AMERICAN NATURALIST [ Von. LIIt

nearer to the ancestors of the ‘‘acridids,’’ although the
line of development of the latter may have branched off
from a ‘‘locustid’’ stock. The Phyllium-like forms seem
to be modified phasmids which have certain features in
common with the grasshopper group.

The plecopteroid, blattoid and orthopteroid groups are
all very primitive, and are so intimately connected by
intermediate forms or synthetic types that they are to
be considered as representing one section of the Ptery-
gota, to which the term ‘‘Plecopteradelphia’’ was ap-
plied (Crampton, 1916a) to indicate that they are the im-
mediate descendants of Plecoptera-like ancestors and the
ephemerids and Odonata should doubtless be included in
the same section of the Pterygota. There is a bare pos-
sibility that the blattoid forms rather than the Plecop-
tera are nearer the ancestral type from which the others
were derived, but the close resemblance of immature Ple-
coptera to lepismids, and the very primitive organization
of the Plecoptera, make it very probable that they, rather .
than the blattids, represent very closely the ancestral
forms which gave rise to the blattids themselves, and the
other types mentioned above. The higher insects were
also apparently descended from forms ultimately de-
rived from ancestors related to the Plecoptera, but they
have ‘‘clustered together’’ in another division forming
the ‘‘Neuropteradelphia’’ (Crampton, 1916a) or forms
grouped about the Neuroptera in the second section of
winged insects next to be considered.

The members of the second section (or ‘‘Neuroptera-
delphia’’) fall into two principal groups. One of these,
comprising the psocids, ‘Thysanoptera, and hemipteroid
forms, were probably descended from ancestors not un-
like the psocids, and it is also quite possible that the Mal-
lophaga, and the Anopleura or ‘‘Siphunculata,’’ rep-
resent offshoots of this stock. This group had a common
origin with the neuropteroid insects (probably from
Plecoptera-like forebears) and the two paths of develop-
ment have extended side by side for a considerable dis-
tance, both having numerous characters in common.

No. 625] THE EVOLUTION OF ARTHROPODS 175

The Neuropteron group comprises the Neuroptera,
Trichoptera, and Mecoptera, with their immediate rela-
tives. They and their descendants are very closely re-
lated to the members of the psocid group mentioned
above, and the two lines soon merge in a common ances-
try when traced back toward the plecopteroid stem. The
Neuroptera seem to be a very ancient type, and have
inherited certain primitive characters which would indi-
cate that their line of development branched off at a com-
paratively low level. Both the Trichoptera and the Me-
coptera are descended from ancestors quite like the pres-
ent-day Neuroptera, while the Lepidoptera branched off
near the trichopteron line of descent, and the Diptera
branched off near the mecopteron line (see also Cramp-
ton, 1917b). The Siphonaptera were apparently de-
scended from ancestors not unlike phorid Diptera.

The Hymenoptera represent a somewhat aberrant
group having affinities with both the members of the
psocid and neuropteron groups. Their line of descent
probably originated near the point at which the psocids
and Neuroptera branched off, and they inherited many
features also present in the members of both of these
groups, so that their line of development must have ac-
companied or extended beside those of the other two for
a considerable distance before it branched off to follow
its own path of specialization.

SUMMARY

The points which should be especially emphasized in
regard to the evolution of the insectan branch of the
arthropod lines of development may be briefly summa-
rized as follows:

The ancestors of arthropods were not of any one type,
but varied in regard to the number of segments compos-
ing their bodies, the outline of the body, ete.; and while
some of them may have been as small as the tardigrades,
it is more probable that the types would be included be-
tween the extremes represented by the Onychophora and

176 THE AMERICAN NATURALIST (Vor. LII

the Annelida, or even between the extremes included
within the mee group itself.

The first arthropods also were not of one single type,
but possibly varied as greatly among themselves as a
branchiopod-like copepod would differ from a copepod-
like branchiopod, ete. It is very probable that the stem
forms eventually giving rise to the line of development
leading up to the production of the insectan type of
arthropod would be included in the branchiopod group.

The next stage in the evolution of the insectan type of
arthropod is represented by forms related to the lepto-
stracan group, although the Leptostraca do not include
all of the types representing this stage of development.
It is possible that the Trilobita may be considered as
somewhat near these forms, since they exhibit a few char-
acters in common with them, but the trilobitan line of
descent is not directly in line with the insectan path of
development, since it diverges toward the evolution of
the merostomes and eurypterids leading off toward the
arachnoid type of development and away from the in-
sectan type.

A further stage of development is represented by the
members of the group including the Anomostraca, Cu-
macea and Tanaidacea. While they doubtless also re-
sembled the other members of this group in certain re-
spects, it is quite possible that the ancestors of insects
and ‘‘myriopods’’ varied between the extremes repre-
sented by Bathynella among the Anomostraca and by
such forms as Apseudes, ete., among the Tanaidacea,
from which the Isopoda, etc., were also descended.
Bathynella, with no eyes, with its cylindrical body, re-
duced legs and ‘‘stumpy’’ pair of pleopods, basal limb
appendages suggesting the precursors of styli, short
terminal appendages, etc., must be very like the ances-
tors of the Protura and Scolopenrelloid forms; while
such Tanaidacea as Apseudes, with its flagelliform termi-
nal uropods, and the type of head appendages, etc., pres-
ent in the Isopoda in general suggest the type of ances-

No. 625] THE EVOLUTION OF ARTHROPODS Fii

tors giving rise to those Apterygota which are provided
with flagelliform terminal appendages.

The members of the Symphyla-Pauropoda group have
retained many characters present in the ancestors of the
‘‘ Myriopoda’”’ and Insecta. The Chilopoda are an off-
shoot from this stock and do not stand quite as near the
direct line of development of the insectan type.

The Apterygota are the nearest known representatives
of the ancestors of winged insects, and while the first
insects to be evolved possibly were of types resembling
both the proturan forms and the campodeoid forms (or
even the machiloid forms), the lepismid type approaches
as nearly as any known forms to the lowest representa-
tives of the Pterygota.

The first winged insects resembled the lepismids in
many respects, and their nearest living representatives
are the ephemerids and Plecoptera. The Plecoptera and
the fossil Paleodictyoptera stand at the base of the lines
of descent of the higher forms, and, since the line of de-
scent of the Plecoptera has accompanied those of the
higher forms for a longer distance, they are even more
important than the Paleodictyoptera for a phylogenetic
study of the evolution of higher insects. Most higher
forms cluster about the Plecoptera and Neuroptera as
nuclei representing synthetic types of the greatest im-
portance, and both types are of considerable antiquity,
although the Neuroptera were possibly ultimately de-
scended from forms not unlike the Plecoptera (and
ephemerids).

It is quite improbable that insects or arthropods in
general (as well as the more inclusive groups) are of a
polyphyletice origin. The ancestors of insects, for exam-
ple, were of several types, some resembling the ancestors
of isopods, while others resembled the ancestors of the
Symphyla, etc., and the lines of development of all three
extend for some distance side by side before each begins
to diverge from the others. Those insects resembling
Symphyla were not descended from symphylid forebears
nor were those insects which resemble isopods descended

178 THE AMERICAN NATURALIST [Vou LII

from isopod forebears, but the symphylid and isopod
characters which appear in certain insects were inherited
from their ultimately common ancestry, and the relative
positions of the different ancestors of insects in the ‘‘he-
reditary areas’’ of this common ancestry (i. e., whether
their hereditary areas were contiguous to those of the
ancestors of isopods or to the ancestors of the Symphyla,
etc.) determines whether certain of the insects descended
from them shall resemble isopods or Symphyla, etc., and
the same principle applies in the successively larger as
well as in the smaller groups of living things.

REFERENCES CITED

Beecher.

1896. The Morphology of Triarthrus. Amer. Jour. Sci., 16, p. 166.
Bernard.

1892. The Apodide.
Boerner i

1902. Arachnologische asi Zool. Anz., 24, p. 433.
1903. Mundgliedmassen E Gi iaekanneacas ’ Siteb. Gesell. Naturf.
Berlin, 1903, p. =
1903. Be Beingliederung der Arthropoden. aoe 1903, p. 58, also
. 292. See also Zool. Anz., 27, p. 226.
1908. Coliembolen aus Sudafrica, ete. L. mo Forschungsreise
westl. u. centri. Sudafrica, etc., IVa, p.
1909. fa Homologien zwischen Crasteecen u. Toe Zool.
Anz., 34, p. 100.
Borradale.
1917. The pag of the Palemonid Prawns. Proc. Zool. Soc.
London, Pt.
Calman.
Crustacea. Lankester’s Treatise on Zoology, Pt. VII.
Carpenter. 3 À
1903. On the pg ai gs between the Classes of Arthropods. Proc.
Roy. Irish Acad., May, 1903, p. 320.
Chamberlin.
916. The Lithobiid Genera Oabius, ete. Bull. Mus. Comp, Zool.
Harvard, 57, 1.
Chappius.
se Bathynella Natans, ete. Zool. Jahrb. Abt. Syst., 40, p. 147.
Crampt
ery oe Sclerites and Systematie Position of Grylloblatta, a
Remarkable Orthopteroid Insect. Ent. N , P 337.
1916a, Lines a Descent of Lower Pterygota. Ent. News, 27, p. 244.
1916b. Orders and Relationships of Apterygotan Insects. Jour. N.
Y. Ent. Soc., 24, p. 267. 5
1916c. Phylogenetic Origin .. . of Wings of Insects. Ibid., p. 1.

No. 625] THE EVOLUTION OF ARTHROPODS 179

1917a, Phylogenetic Study of Head, Neck and Prothorax of Aptery-
gota and Pterygota. Ent. News, 28, p. 398.
1917b. Phylogenetic Study of Head in Neuroptera, Mecoptera, Dip-
tera and Trichoptera. Ann. Ent. Soc. America, 10, p. 337.
19184. pectin Study of ‘Terminal Abdontinal Structures or
erygota and Pterygota. Jour. N. Y. Ent. Soc., 26, p.
1918b. Pakea Study of Terga and Wing Bases of Plecoptera
Embiids, Dermaptera and Coleoptera. Psyche, 24, p. 7
Handlirsch.
1903. Zur Phylogenie der Hexapoden. Sitzb. K.. Akad. Bethe Wien,
1903, 112, p. 716. See also Zool. Anz., 1904, p.
1906. Phylogenie der Arthropoden. Verh. Zool. Bot. Fresi Wien,

» Dp. 88.
1906. Die Fossilen Insekten.
Heymons.
1901. Entwicklungsgeschichte der Scolopender. Zoologica, 13, Heft
33, p. 1.

Jaeckel.
1901. Beitr. z. Beurtheilung der Trilobiten. Zeit. deuts, geol. Gesell.,
53, p

Lankester.
1872. Remarks on Pedalion. Q. J. M. S., 12, p. 338.

Lignau.
1911. Entwicklung des Polydesmus. Zool. Anz., 37, p. See also
í Mem. Soc. Nat. Odessa, 38, p.

Lindstroem.
1901. Visual Organs of Trilobites. Kgl. svensk. Vet. Akad. Handl.,

, P
Il.
1917. Organic Evolution,

en.

1912. Evolution of Vertebrates.
Philiptschenko.,

1912. Beitr. z. K. der Apterygoten. Zeit, Wiss. Zool., 103, p. 519.
Ruedemann.

1916. Presence of Median Eye in Trilobites. Proc. Nat. Acad. Sci.,

dabinshart.
1915. Text-Book of Geology.
Silvestri.
1913. Descrizione di un nuovo ordine di insetti. Boll. Lab. Zool. gen.
agr. Portici, 7, p.
Tothill.
1916. Ancestry of Insects. Amer. Jour. Sci., 42, p. 373.

1911. C Cambrian RATOA Skt, Smith. Mise. Colls., 57, No. 2,

P.
1912, e Cambrian Branchiopoda, Malacostraca, Trilobita and
Merostomata. JIbid., No. 6, p. 148.

914. On Grylloblatta campodeiformis. Can. Ent., 46, p. 93.

SHORTER ARTICLES AND DISCUSSION

ON THE pap OF FUNDULUS TO CONCEN-
ATED SEA WATER!

I. THERE is at Bermuda a Fundulus, decoribed by Giinther
(79) under the name F. bermuda, which is very closely related
to F. heteroclitus, if not indeed specifically identical with it.’
The common habitat of this Fundulus is along the shores of
mangrove swamps, in water normally having a salinity of 35-36
per mille (Cl—20+ per mille; sp. gr. about 1.0225°%°).
When this Fundulus was placed in sea water which was allowed
to evaporate at laboratory temperature (about 27°) a good
number of specimens were found to resist a concentration of
about 154 M sea water (Cl= 67 per mille). According to Loeb
(713, 716), F. heteroclitus at Woods Hole may be brought to live
in a concentration equivalent to 1% M or 1% M, if the water be
slowly evaporated, but a 1% M concentration is rapidly fatal.
Sea water at Woods Hole is at about M/2 (salinity 32 per
mille +), with a freezing-point depression of 1.81° (Scott, ’13),
whereas the Bermuda sea water is nearly %¢ M, with (accord-
ing to Knudsen’s Table, 5) a freezing-point depression of 1.95.°
McClendon (’11) found the A of Tortugas water (S —36 per
mille +) to be 2.03°.

Is the considerable difference noted in the resistance of Fun-
dulus taken from these differing environments to be regarded
as an instance of adaptation brought about in nature?

II. Tests were made at different seasons to discover the upper
limit of concentration which the Bermuda fundulus will toler-
ate. One of these experiments may be cited as an example:

Experiment 4.—Aug. 20, 1917. Six fundulus were placed in each of
three glass aquaria containing 2 liters of sea water (Cl==19.65°/,,3;
S = 35.50°/,,) brought from the mangrove creek in which the fundulus
were collected. The water was allowed to evaporate at room tempera-
ture (28°). In aquarium No. 1, two fishes were still alive on Sept. 7,

1 ea from the Bermuda Biological Station for Research, No. 102.

2Some of the specimens used in these experiments were examined by Mr.
Samuel aay. of the Museum of Comparative gh me who nee
them to be Fundulus heteroclitus, var. bermude Goode

180

No. 625] SHORTER ARTICLES AND DISCUSSION 181

at which time the water was so concentrated that the salinity titration
(of an aliquot part of a dilution with distilled water) gave Cl=
66.69°/,,. These two fishes lived until Sept. 11, when the Cl content
of the water was 72°/,,. Similarly, in the other two aquaria the max-
imal concentrations were Cl= 66.83, 66.90°/,,.

Other tests gave comparable results, about one third of the
fundulus living until the Cl content of the water was nearly 67
per mille (= 154 M sea water). Provided the process of evapo-
ration occupied at least two weeks, slowing the evaporation of
the water did not seem to augment the resistance of the fishes.
Fundulus ‘‘adapted’’ by slow evaporation lived in 14% to 15% M
solutions for a week or more. When a concentration of about
Cl==70 per mille (1% M) was reached, the fishes usually died
very rapidly, although an occasional one survived until the Cl
content was 75 per mille, whereas at Woods Hole, according to
Loeb, the rapidly fatal concentration is 124 M. In each case the
maximal concentration endured for any length of time is about
three times that normally experienced by the fish.

III. While Loeb found the resistance of F. heteroclitus to be
only slightly enhanced by a series of ‘‘adapting’’ experiences in
waters of gradually increasing concentration, it might neverthe-
less be argued that a more gradual series of changes, leading to
normal life in more saline water, would be more efficacious.
_ There are several facts which dispose of this supposition, aside
from the somewhat disproportionately great increase in abso-
lute resistance which is exhibited by the Bermuda form.

The Bermuda fundulus is found not only in the mangrove
creeks, but also in certain landlocked brackish ponds (e. g.,
Warwick Pond, Trott’s Pond) where the salinity is usually
14.5-23 per mille (Cl=8.0-12.7 per mille), although it varies
somewhat with the rainfall. The level in these ponds rises and
falls slightly with the ocean tide, but there is no variation in
salinity synchronous with this. Fundulus were taken from
these ponds and placed in sea water which was allowed to evapo-
rate slowly, and others were put in pond water which was al-
lowed to evaporate. The lethal concentrations in these two
Series were practically identical, namely, at about Cl = 65-70
per mille, one third of the individuals usually surviving until
the Cl content reached 66-67 per mille, which is essentially the
same maximal concentration as that found with the individuals
living normally in undiluted sea water. F. bermude will live

182 THE AMERICAN NATURALIST [Vor. LIII

for a long time in rain water containing but a trace of salts, and
those from the brackish ponds will live equally well when sud-
denly transferred to sea water of full salinity (36 per mille).

ow, the fundulus living in the brackish ponds have been
there for an indefinitely long period. They reproduce there,
and must be regarded as ‘‘adapted’’ to the low salinity of the
ponds. There is consequently no reason to expect, on the
adaptation hypothesis, that they should be as resistant to con-
centrated sea water as the individuals living in Fairyland Creek,
for example, where the water is of normal salinity. Yet this
appears to be the case. It is true that this species inhabits other
brackish swamp pools at Bermuda, where the salinity under-
goes considerable changes. But if the high resistance of the
isolated-pond fundulus were to be explained as the result of a
persisting mechanism inherited from ancestors adapted to with-
stand changes in salinity, then it will be noted that the appeal
to adaptation in the first place becomes not merely superfluous,
but inconsistent.

IV. There is another explanation available, which probably
accounts for the high resistance of the sea-water and brackish-
pond fundulus to concentrated solutions. This explanation con-
siders that the conditions of temperature and the composition
of the water (especially in the brackish ponds) have shifted the
protoplasmic equilibria which determine the composition (and
hence the permeability and the eo of the limiting mem-
branes of the fish’s body.

Loeb and Wasteneys (’12, ’15) found that fundulus taken
from a temperature of 10° died in the course of several hours
when kept at 29°, in a few minutes at 35°; whereas those main- `
tained at 27° would live indefinitely if transferred to 35°. The
fundulus at Bermuda living in the mangrove creeks are at a
temperature of 26°-27° (during the summer months). They
withstood for some hours a temperature of at least 37°, and died
when heated to 40.9°. In the shallow landlocked ponds the sur-
face temperature was 30°-33°. Fundulus from these ponds
withstood for several hours a temperature of 39°-40°, and died
when heated to 42.6°. The upper temperature limit was also
determined for fundulus from the brackish ponds which (at
27°) had for two weeks been living in 1°54 M sea water; they
withstood 40°, and died at 42.5°. Other individuals living for
the same period in % M sea water withstood 40°, and died quickly

No. 625] SHORTER ARTICLES AND DISCUSSION 183

at 41.5°. There is thus seen in fundulus the correspondence
usually found in every thermal species between the temperature
at which the animal lives and the maximal temperature which
it can successfully withstand (cf. Mayer, ’14).

The alkalinity of the waters inhabited by the Bermuda fun-
dulus is quite various. In the mangrove creeks the reaction of
the water along the shore may vary a little with the state of the
tide, but is usually not far from py=8.1. In the landlocked
brackish ponds, however, the alkalinity is commonly much higher
than this. In one pond, where many alge were growing, the alka-
linity was conspicuously high, py,—9.0-9.2 (except after
rains) ; and in another, with a sparser growth of water plants,
the reaction usually observed was py==8.7. Rain water had at
this time a consistent reaction of p,—5.9-6.0, but after contact
with the soil and limestone it quickly becomes alkaline, so that
the water in cave pools, or dripping from growing stalactites,
was found to have a reaction of p= 7.9-8.0. The high alka-
linity of the pond waters may be important in determining the
survival of fundulus in abnormal solutions.

If the idea is correct that the composition (e. g., the calcium
content) and (?) the temperature of the sea water or pond
water are responsible for the high resistance of the brackish-
pond fundulus to concentrated sea water, then we should expect
that NaCl solutions would be less toxic for the Bermuda fundu-
lus than for the northern variety, which, according to Loeb, and
Wasteneys (12), is killed by 1 M NaCl in less than one half
hour (at about 18°-20°, it is inferred). The pure NaCl solu-
tion increases the permeability of the surface membranes of
fundulus. At 25°-27°, 50 per cent. of the Bermuda fundulus
lived forty-five minutes in 1 M NaCl solution, when the speci-
mens were taken from the mangrove creeks. Individuals from
the landlocked brackish ponds lived about the same length of
time (even after rapid washing, three times, in changes of NaCl
solution). At 20° they lived a little longer. This result is in
agreement with that obtained experimentally by Loeb (16, p.
332), namely, that fundulus adapted to higher concentrations of
sea water became more resistant to pure NaCl solutions; those
brought artificially to live in 1% M sea water could live two to

3 Moore and his co-workers (’14) state that the photosynthetic activities
of alge are capable of increasing the alkalinity of sea water until py = 9.0;
I have been able to confirm this statement through experiments with the
green alga Valonia.

184 THE AMERICAN NATURALIST [Vou. LIII

three days in % M NaCl, which killed fundulus taken directly
from the sea in less than four hours. But in the present case
the fundulus from the brackish ponds lived about equally well
in pure NaCl solution. The cause of this behavior, which would
not be expected from an adaptational standpoint, is believed to
lie in the direct effect of the calcium or some other element of
the pond water. In % M NaCl solution fundulus from the man-
grove creeks and those from a brackish pond lived respectively
3.5 and 4.0 hours, roughly, at 20°. The experiments with 1 M
NaCl seemed more valuable for the purposes of this inquiry,
because of the more rapid toxic effects, secondary complications
being thus more easily avoided.

V. The fact that the Bermuda fundulus, closely related to F.
heteroclitus, but living usually in water of greater salinity than
that inhabited by the Woods Hole variety, seems also able to
withstand a distinctly higher concentration of evaporated sea
water than the latter will tolerate, is therefore not to be con-
sidered an expression of adaptation to life in more saline water.
Other members of the same species at Bermuda which are con-
fined to brackish ponds of low salinity, and have for at least
several generations been restricted to this environment, are
equally resistant to concentrated sea water, and to pure NaCl
solutions, and more resistant than the fundulus at Woods Hole,
indicating that the resistance of the Bermuda form is due to a
direct action of certain constituents of the waters in which it
lives upon the composition of its surface membranes.*

REFERENCES
Giinther, A.
1879. Rept. Voy. Challenger, Zool., Vol. 1, Pt. 6.
Loeb, J.

1913. Biochem. Zeits., Bd. LIII, p.
1916. ‘The Oranii as a Whole,’’ anne York; x + 369 [pp. 328
et seq.].
Seott, G. G.
1913. Ann. N. Y. Acad. Soi, Vol. XXIII, p. 1.
observations here briefly reported in a preliminary way have been
temporarily interrupted, but it is hoped to continue them in the near future,
indication was noted of a lowered resistance to pure NaCl corre-
lated with a decreased alkalinity of the pond-waters, during the winter
months. The determinations of alkalinity were made with the aid of ap-
paratus purchased through a grant to the Station from the C. M. Warren
Fund of the American Academy of Arts and Sciences.

No. 625] SHORTER ARTICLES AND DISCUSSION 185

Knudsen, M.
1903. Publ. de Circon., No, 5, Cons, Perm, Intern. Expl. Mer.
McClendon, J. F.
1911. Carneg. Instn. Wash., Year Book No. 9, p. 128.
Loeb, J., and Wasteneys, H.
1912, Jour. Exp. Zool, Vol. 12, p. 543.
1915. Jour. Biol. Chem., Vol. 21, p. 223.
Mayer, A. G.
Carneg. Instn. Wash., Publn. 183, p. 1.
Moore, B., Prideaux, E. B. R., and Herdman, G. A.
1914. Trans. Liverpool Biol. Soc., Vol, 29, 233-264,
W. J. CROZIER
PEMBROKE,
BERMUDA, January, 1918.

A NOTE ON THE FATE OF INDIVIDUALS HOMO-
ZYGOUS FOR CERTAIN COLOR FACTORS
IN MICE

ISBEN and Steigleder have reported on certain breeding ex-
periments with mice which produce evidence in support of the
view advanced by Castle and the writer in 1910, and later
strengthened by Kirkham, 1917, that homozygous yellow mice
were formed but perished dring embryonic life.

At the time that they were collecting their data, the writer
was, on a smaller scale, carrying on similar experiments. In
the course of these experiments, certain data confirmatory to
the results of Isben and Steigleder and of Kirkham were ob-
tained. It seems best at this time to put these results on record.

The embryos referred to as ‘‘abnormal’’ may be considered
as falling in Isben’s and Steigleder’s Class A of dead embryos,
that is to say, those in which development ceased shortly after
implantation as contrasted with those in which death had re-
sulted probably from overcrowding within the uterus during
the latter part of the period of gestation.

Three types of matings to control the results in yellow X
yellow crosses were made. In all cases, the non-yellow animals
used were taken from the same stock as that producing the yel-
lows. The control matings made were as follows: Yellow female
X non-yellow male, non-yellow female X yellow male and finally
non-yellows crossed inter se. The numbers obtained are small
and are grouped together in the following table:

186 THE AMERICAN NATURALIST [Von. LIII

TABLE I
Q g Normal Abnormal

11696 Brown < 11713 Yellow ........... Ha pna
11716 Browh: X IL7IT Yellow i iss ok, 7 =
11438. Brown- Brown: iriser ari ese ee 7
11442 Yellow X Broni zeier. Seite. 6
11562 Yellow X Brow 00 is at Os 9
10619 Yellow X Brown or Black ......... 2

42 T

The one abnormal embryo consisted of a small apparently
embryonic mass, with a blood clot closely jammed in between
two normal embryos. It will be noted that from these matings
97.6 per cent. of the embryos are normal and 2.4 per cent.
abnormal.

When yellows are crossed inter se a very different result is
obtained, as may be seen from the following table, which shows
the result of such matings:

TABLE II
ç g Normal Abnormal

‘COHOW DK SOOW i osia eee ee een 4 1
Wellow H Yow . Posi eee, Aa lerde ci 3
VYollowal Yellows. saves 2 12s eee. 7 2
Yelow GSO VOW sik oan wie Ode aie wits HOR ENE i 3
1867 WELEI eki atures wets ges 6 0
11786 POUOW isikae s wens ee ake melee 8 2
KONOW DPC EOUOW oe rove eee hou eee ea ss 7 1
Mellow BS Vellow. a0 tis et we wee 10 0
11151 SOURIS aa TE waded cu A O an oi 2 1
12916 K YUOA oroi is ce iaa 5 0
Xow d X YOUOW aiioe osas Eana vi 3
1 W Yellow Seek rs ee lie cen 4 2
Yellow A NX Yollow i 0 Se eek aes 7 0
Tahe OS ollow is neces. caves es oe 4 1
1926 X Geet eran sae ake te eee 3 0
12672 X — 99 Sooty Yellow ........... 4 2
91 21

18.7 Per cent. abnormal,
81.3 Per cent. normal.

From this table it will be seen that 81.3 per cent. of the em-
bryos produced are normal, and 18.7 per cent. are abnormal.
If one considers in addition the fact that Kirkham obtained
embryological evidence that certain embryos broke down even
before implantation, it seems probable that the fate of the homo-
zygous ‘gation mouse is known.

No. 625] SHORTER ARTICLES AND DISCUSSION 187

One other point of some interest should be noted. In 1915
the writer reported on the hereditary behavior of black-eyed
white spotting in mice. At that time it was found that this
character behaved in a similar manner to yellow in that no
animal homozygous for it was obtained. Later it was found
(1917) that black-eyed white spotting was, however, entirely in-
dependent of yellow in heredity, although its behavior was
analogous.

If the uteri of black-eyed white females which are pregnant
by black-eyed white males are examined they are, in some cases,
found to-contain a certain number of abnormal embryos of the
same gross appearance as those occurring in the yellow X yellow
matings. The numbers obtained are small but striking.

TABLE III
Normal Abnormal
Black -eyed White A x Black -eyed White ...+..<05++ 3 1
Cipher oes s 1 5
Black -eyed White B X iak eyed White ..........- 5 0
1 Blackeyed White ........... 7 0
xcs 6

The percentage of abnormal embryos is 27.2. While this last
mentioned cross should be repeated, it nevertheless indicates
that, like the homozygous yellow embryo, the homozygous black-
eyed white embryo breaks down, in most cases at least, after its
implantation in the uterus.

C. C. LirtLe
December 2, 1918

LITERATURE CITED

Castle, W. E., and Little, C. C.

1910. Science, N. S., Vol. 32, pp. 868-870.
ee H. L, and Gtstyieder: E.

917. Am < Nat, VoL 5i, pp. 740-752.

ides W. B.

1917. Anat. Record, Vol. 11, pp. 480—481.
Little, C. C.

1915. AM. Nar., Vol. 49, pp. 727-740.

1917. Genetics, Vol. 2, pp. 433-444.

188 THE AMERICAN NATURALIST (Von. LIII

THE VARIETIES OF HELIANTHUS TUBEROSUS

THE girasole, Jerusalem artichoke or sunroot, Helianthus
tuberosus of Linnæus, has been in cultivation more than three
hundred years. It is native in North America, and its tubers
were well known as a source of food to the Indians in pre-
Columbian times. In spite of its long history and value as a
‘*root-crop,’’ this plant has received little attention from breed-
ers in modern times, and it still remains to be seen what may
be done with it, with intensive study and improved methods.
At the present time we can say that it is enormously prolific,
and the tubers are excellent food for man and beast. Recent
experiments indicate that they may be an important source of
sugar in the form of syrup. The very large tops can be’ used
as fodder. For these and other reasons it is desirable to inves-
tigate the existing varieties, and place on record their principal
characteristics. This year, in Boulder, Colorado, I have grown
all those listed below, excepting the first:

(a) typicus.—I take as typical of the original H. tuberosus the Sait
figured by Fabius Columna in his account of little-known and rare plants,
published in 1616. This figure is cited by Linneus. It is labelled Flos
Solis Farnesianus, Aster Peruanus tuberosus, i. e., the Farnesian sunflower,
or tuberous Peruvian aster. It did not, of course, originate in Peru. The
figure shows that the plant was much branched, the branches highly fl
iferous; tubers quite large, potato-shaped or oblong; leaves short- AA
with broad base, the margin quite coarsely crenate- ale, rays about 16,
not very long; involucral bracts recurved. T have never seen a plant with
AeA this combination of characters, but the peculiarities recur separately

n different varieties.

(b) nebrascensis.—Received from the Rev. J. M. Bates, who found it
growing wild at Red Cloud, Nebraska. It is like typicus in Be general
appearance im many floriferous branches. Compared with "i
scribed below) it differs conspicuously by the shiny upper surface o. pe
and the less densely hairy stems. It flowers earlier than the cae
forms with large tubers. The heads in bud have the involucral bracts
spreading (as in typicus), dark basally, much less hirsute than in albus.

les are much longer than in the varieties with s tubers, their

ase about 42 mm, (30 or less in the large-tubered forms), so the flowers

are very handsome. The tubers are produced at the ends of the rhizomes,

mostly distant from the stem, and are elongate, broad or shea whee

ut usually not claviform, and not compressed at end. The thin skin is
pale =e wn.

(c) alexandri.—Growing wild in Michigan, and received from the late Mr.
S. Alexander, who regarded it as a distinet species. It resembles the tall
cultivated forms in not being P ae branched or bushy, as are the

No. 625] SHORTER ARTICLES AND DISCUSSION 189

two varieties described above. Compared with albus it differs by the
opposite leaves, less densely hairy stem, bases of leaf-blades more abruptly
truncate, yet upper part of petiole much more broadly winged; leaves longer
in proportion to breadth. hairs on midrib beneath subappressed (erect in
albus). The upper surface of leaves is dull, as in albus. The ligules are
long, as in nebrascensis, and are not rarely quilled. The stigmas begin to
emerge while the anthers are fully extended, which is not the case with the
other forms. The tubers are elongate, at the ends of the rhizomes, clavi-

form, pied drical, ess compressed apically. They are white,
ith y thin wish skin, the color eo like that of
d) imide eau out by the firm ohn Lewis Childs as ‘‘ Pink

Helianthi.’’ A request for information Sieve its origin brought no
answer. It is a small-tubered Bes rge ably ae in the state in which
it occurs wild. The mode of gr nd general appearance are as in
nebrascensis, but the leaves are ae dull ss The large leaves are
coarsely dentate, with ve Seg base, but the mor is not so broadly
winged apically as in PEA The rays are sp in nebrascensi.
This is e first of the a. to come into flower; one pa ad was o
August 21. The tubers are comparatively short, usitorm, Pinares, not
much attenuate at ends; they are about 50-70 mm. long a
diameter, produced at rian of rhizomes. The color is as pinkish-purple,
as in variety purpur
/ (e) fusiformis. gen ‘*Rose’’ variety of Sutton and Sons, Reading, Eng-
and. We are indebted to the Sutton firm for y eee us with
ae of their cultivated varieties. This is a kable form, very
distinct from all the S rg it first comes ai ø pies slowly, and
tends to spread out on the grou When mature it is about 7 feet high,
only about two- ae a height a albus and purpureus. The stems are
entirely green, not purple above as in albus. The leaves turn yellow in the
r without any of the red so conspicuous in albus. After frost, most of
stems give way somewhere above the ewes soi the ih above hangs
rt forming an acute angle with tanding s This rarely
occurs in albus, but not in purpureus, nor in Pn per a Te leaves are
long, with a cuneate base, which is very distinctive., The margin is irregu-
larly dentate. The involucral bracts are paler and much longer es in
albus. The plants were just coming into flower September 22, are
later than any of the other forms. The tubers are large, of ae ie
but more or less fusiform, with only —= lateral knobs. The diameter
is about 45 mm., the length two or three tim uch. The se is pale
brown, Benctionliy the color of petals ene a faint rosy suffusion.
The tubers of one plant weighed 8 Ibs
v (f) albus—We first got this, a number of years ago, from Dreer of Phila-
delphia. Mr. L. Sutton tells me that his firm first offered it in 1915, having
obtained it from some one who said it had been sent him by a friend in
South Amreica. He believes it had not been grown in England before this.
Dreer had it much earlier in this country, having obtained it from Mr. A. E.
Coleman of Enonville, Va. Mr. Coleman states that he knows nothing of
the origin of the variety, and hardly thinks any record wa This
variety is very tall, and usually not very conspicuously ‘aude: The upper

190 THE AMERICAN NATURALIST [Vou. LIII

Fig, 1. Helianthus tuberosus var. fusiformis at Boulder.

part of the stem is purple, and in the fall the upper leaves turn very red.
The leaves have the blades nad A or subeuneate at base, the larger
leaves forming an angle greater than a right angle; the petioles are not
broadly winged apically. ine this differs greatly from the wild alexandri,
and in addition the margins are sharply though rather finely dentate, while
in alexandri they are crenate. The axillary branches have a purplish-black
callus at base above; in nebrascensis this callus is reddish. The areg
are conspicuously longer than in nebrascensis. The heads in bud have
phyllaries or involucral bracts erect, not spreading as in nebrascensis ho
typicus. The ligules are about 30 mm. long and o
purpureus are considerably broader, 30 mm. long and 11 Droë: The tubers
re ve arge and knobby, irregularly subglobose, and mostly among the
roots, close to the base of the stem. One plant of Sutton’s white, dug Nov.

had 12 lbs. of tubers. The color of the tubers is white.

A subvariety of albus, with more deeply serrate leaves, was kindly sent
si Mr. G. C. Worthen, who purchased the tubers in Boston. The growth

other characters do not differ, and the tubers are the same, the buds

rplish.
g) purpureus.—Received from Sutton, who states that it is the jesa
long cultivated in England. It is a tall plant, with the appe
and manner of growth as albus. On June 30 I noted ak as DEN with

albus it had paler, larger eaves and the veins were more impresse I
was in good flower by Sept. 22. The si area are notably piedi or
deflexed; the ligules are bradi than in clbus, Both purpureus and fusi-

formis have an orange flush at the base of the ligules, which is lacking in
albus. The ends of the disc-bracts are broader and more hairy in fusi-

No. 625] SHORTER ARTICLES AND DISCUSSION 191

Fic. 2. Helianthus tuberosus var. albus (var. fusiformis at extreme left). The
owers are Helianthus annuus.

formis than in albus; in purpureus they are. much as in fusiformis, but the
difference from albus is ha rdly so marked. The stems show no red color.
The tubers are like those of albus, but are rosy-purple, the same color as
those of purpurellus. ` One plant produced seven lbs. of tuber

It will be seen from the above, that all the varieties differ in
a number of characteristics. At the same time, they agree in
various particulars. Thus purpureus and purpurellus in the
color of the tubers, purpureus and albus in their shape. We
do not know how far the cultivated varieties owe their charac-
ters to aboriginal ancestors; but it is practically certain that no
wild form has tubers as large as those of the cultivated ones.*
It is also certain that the excellent (from our standpoint) char-
acter of having the tubers clustered about the crown, making
them easy to harvest, could not have existed in a wild ancestor,
in which it would be extremely detrimental. On the other hand,
the tubers of albus and purpureus are very knobby, and so hard
to prepare for the table; those of the wild forms are essentially
Smooth (like a sweet potato), but too small. The variety fusi-
formis combines large tubers with, at least in large measure,

1 There is some reason for thinking that the Indians had a cultivated form
with rather large tubers.

192 THE AMERICAN NATURALIST [Von. LIIN

the better shape of the wild varieties. If purpurellus, shaped
like a Zeppelin, could be crossed with another form to secure a
large tuber while conserving the form, the result would be valu-
able. It still remains to determine the chemical constituents of
the several varieties, and this will be done during the winter.

From the standpoint of genetics, an interesting feature is the
distribution of the anthocyanin pigments. - The variety pur-
pureus, with a great quantity of anthrocyanin in the skin of
the tubers, lacks this coloring in the leaves and stems. The
variety albus has it in the leaves and stems, but not in the
tubers. The physiological significance of this is at present un-
explained.

One of the greatest difficulties in the way of plant breeding
comes from the impossibility, in so many cases, of making sure
of the history or oven the identity of the varieties used. The
same thing may go under several names, or the same name may
be applied to different things. In the case of species, it is usu-
ally possible to unravel the synonymy by reference to the origi-
nal descriptions, or to refer to the type specimens. With hor-
ticultural varieties, there is usually no type and no formal
description. The history, in the majority of cases, is lost. When
a new variety is introduced, the firm putting it on the market
rarely states where it came from, and often, after a few years,
ean not recollect. There is no way to ascertain definitely that
what is sold today under a certain name is identical with the
plant bearing that name a number of years ago. These condi-
tions lead to many misunderstandings and difficulties of all
sorts, and to much waste of time and energy. They are no
longer tolerable, when the production of new plants is of such
prime importance to mankind. What we need is an organiza-
tion or office, with suitable means of publication, to study and
report on every plant put upon the market as new. Each should
be carefully described in botanical language, and if necessary |
figured. Its origin, if ascertainable, should be precisely stated,
with full details. Any firm refusing to submit its alleged nov-
elties to such a test, and to permit the reports to be made, would
be under grave suspicion of fraud. Not only would plant breed- -
ers be greatly benefited, but the general plant-buying public
would be saved enough useless expense and annoyance to much
more than pay the cost of the undertaking.

T. D. A. CocKERELL
UNIVERSITY OF COLORADO

THE
AMERICAN NATURALIST

Vou. LIII. May-June, 1919 No. 626

ADAPTATION AND THE PROBLEM OF “ORGANIC
PURPOSEFULNESS’’?

DR. FRANCIS B. SUMNER

Scripps INSTITUTION FOR BIOLOGICAL RESEARCH, La JOLLA, CALIF.

I. Tue REALITY OF THE PROBLEM

Despite the ‘‘revolutions of thought,’’ which succeed
one another with rather bewildering rapidity these days,
we may occasionally listen with profit to the voice of a
past generation. And I can not believe that we are yet
in a position to wholly reject Herbert Spencer’s well-
known characterization of life as a ‘‘continuous adjust-
ment of internal relations to external relations.” Now
it is this process of adjustment to which we give the name
adaptation, and the special structures or functions by
which the adjustments are carried out are called adapta-
tions.

By earlier biologists and philosophers these facts of
adaptation and adaptedness were regarded as among the
most fundamental phenomena of life. It was facts such
as these that furnished ammunition for Paley and a
whole succession of natural theologians. It was these
which Lamarck sought to explain by his theory of evolu-
tion through functional activity, and which Darwin at-
tributed to the action of natural selection. And it is in
this same realm of facts that the vitalists find their rea-

1In its main outlines, this paper was written about five years ago. It was
submitted for publication February, 1918, and since then has undergone
relatively little revision. For this reason, adequate reference has not been
made to certain recent papers.

193

194 THE AMERICAN NATURALIST [Vot LOI

sons for attempting to remove biology from its place
among the natural sciences.

Now it is a curious circumstancè that recent develop-
ments of the evolution theory have carried us continually
‘farther from an explanation of adaptation. Natural se-
lection, in the Darwinian sense, has been relegated to a
secondary position, while the Lamarckian principle is
denied in toto by many. On the other hand, all that the
mutationist can tell us in regard to the matter is that
such useful characters as spring full-fledged into exist-
ence are not likely to be eliminated. Thereupon, the vi-
talist takes fresh hope and asserts the inadequacy of
what he calls ‘‘mechanistic’’ biology to account for pro-
gressive evolution.

Of course, one way to solve a problem is to deny that
the problem exists. And this is what is being done by
various persons who are interested in minimizing the
difference between the living and the non-living. Thus
one physiological botanist, Livingston,’ tells us that any-
thing organic or inorganic is adapted to do just those
things which in reality it is found to do. And he seems
to think it quite as reasonable to speak of the adaptation
of fragments of pumice to float on water, as of the adap-
tation of a flower to insure the visits of insects. The
‘‘concept of purposeful adaptation,’’ which still plays
such an extensive rôle in biology, is due to the fact that
‘fours is a developmentally young science,’’ retaining
‘‘features of its early youth.’’ Sooner or later this con-
cept will be ‘‘totally abandoned, even as the same
concept has already been abandoned by the other natural
sciences. ’’

Now it may well be that a more mature state of science
will enable us to dispense with such naive expressions as
imply that an organ has a function to perform in the
economy of an animal. And it may be that growing en-
lightenment will lead us to replace such a primitive no-
tion as that of life by chemical affinities, electric charges
and what not. But until that happy (or unhappy!) day

1* AMERICAN NATURALIST, January, 1913.

No. 626] ADAPTATION 195

arrives, I think that most biologists will continue to re-
gard the origin of adaptive characters in animals and
plants as offering a real and important problem for
solution.

Again, Parker,? while going to no such lengths as this
in denying the significance of organic adaptation, ex-
presses his belief that
the majority of animal reactions are, in all probability, neither con-
spicuously advantageous nor disadvantageous to the life of the indi-
vidual. They are dependent chiefly on the material composition of
the given organism, and, so long as they are relatively indifferent to the
continuance of life, they pass without special consequence. ... The
world at large affords an environment in which each animal has a wide
range for possible reactions, and of a number of responses that might
be made to a given set of conditions, one may be quite as appropriate
for the continuance of life as another. In other words, versatility
seems to be a more truthful description of actual conditions in animal
life than the rather rigid state implied in the idea of adaptive responses.

It may be freely granted that much ingenuity has been
displayed in discovering adaptations which probably do
not exist. And it is doubtless true that many organisms
may live indifferently under a wide range of conditions,
or may eat indifferently a wide range of foods. But is
not this condition of versatility itself a fact of adapta-
tion? The ‘‘continued adjustment of internal relations
to external relations’’ implies that the external relations
change. Man, it is true, may live indifferently on the
equator or within the Arctic circle. But would any one
maintain that the physiological states which adapted him
to these unlike conditions did not differ widely in the two
localities? I may make a meal equally well of meat or
of vegetables. But the digestive fluids secreted for the
occasion would differ in the two cases.

gain, because two wholly unlike plants grow side by
side in the same soil, it does not follow that they are ad-
justed in quite diverse ways to the same set of conditions.
The environment of an organism doubtless comprises
the totality of things which surround it. But the effec-
tive environment comprises only those things with which

? AMERICAN NATURALIST, January, 1913.

196 THE AMERICAN NATURALIST [Vou. LIII

the organism comes into functional relation. And as
the organism evolves, these effective elements become
different.

As I see the situation at present, the fact of organic
adaptation remains the central one in evolution, and in-
deed the central one in biology. I shall give no further
time, therefore, to justifying a rather laborious attempt
to show how this fact may be accounted for without car-
rying us outside the limits of natural science. Before
passing on to this discussion, I will merely remark that I
place in the category of adaptation anything which in-
creases the adjustment of the organism to the conditions
of its existence, whether or not this may ever have a de-
termining influence in the preservation of life. Many
such adjustments have arisen in our own race which cer-
tainly can have played no part in the survival of the indi-
vidual or the race. An example of this is the case dwelt
upon by Spencer, of the correspondence between the
nicety of tactile discrimination on various parts of our
skin and the relative frequency of contact with foreign
objects on these surfaces. And such cases could be mul-
tiplied indefinitely. Nevertheless, much recent biolog-
ical speculation has been vitiated by the identification of
adaptation with self-preservation.

II. ADAPTATION AND ‘‘CoNTINGENCY’’

If an intelligent animal is confronted with the neces-
sity of taking action to avoid injury or secure food, two
ways only would seem to be open to it:

1. It may consciously adapt its actions to this end, or

2. It may go through a series of more or less random
movements until it happens to make one which is fitted to
the needs of the situation.

On first thought, it might seem that these two modes of
procedure were radically distinct, and indeed, in a sense
they are. Considered historically, however, the second
may be regarded as a step in the development of the first,
or, to express the same thought otherwise, intelligent
action is in every case the outcome of earlier experi-

No. 626] ADAPTATION 197

mentation. We can foresee the results of an action, only
in so far as they have been experienced before, either in
a situation identical with the present one, or at least in
situations having certain elements in common with it.
Furthermore, in the early life of the individual, the
movements of the so-called ‘‘voluntary’’ muscles were
in a high degree random and undirected. The associa-
tion between a given muscular contraction and a given
result in consciousness must, in the first instance, have
been purely arbitrary, and could not have been antici-
pated prior to experience.

Thus to restate somewhat paradoxically our original
proposition, an intelligent animal attains a sought-for
end, either by blundering into it or by directing its course
on the basis of past blunders. In either case, the asso-
ciation between the means employed and the end attained
is, in the last resort, accidental. At the outset, the idea of
the end did not in any direct way call forth the means to
its realization, however purposive the action may appear
when fully perfected.

Let us extend our argument to those fields of organic
activity from which intelligence seems to be largely or
wholly excluded. In instinctive actions, even more than
in intelligent ones, a series of movements proceeds unfal-
teringly to a given end, as if directed by the latter. In
earlier days the adaptive instincts of certain lower ani-
mals furnished some of the most telling arguments for
the special interposition of an all-wise Providence. To-
day, as biologists, we commonly explain these movements
on the basis of an inherited ‘‘mechanism.’’ We may be-
lieve, with Loeb and others, that we have to do with a
chain of reflexes, each serving as a stimulus to call forth
its successor at the appropriate moment.

How this mechanism arose is a disputed point, but
there are two principal hypotheses as to its origin: (1)
Instinctive actions are ones which originated, intelli-
gently or otherwise, in the course of individual expe-
rience, and finally became fixed through heredity; and

198 THE AMERICAN NATURALIST [Vou. LII

(2) they are the result of natural selection, acting on
congenital tendencies toward such movements as proved
to be adaptive.

Without discussing the merits of these rival theories,
which are by no means mutually exclusive, I merely wish
to point out that both of them assume a complete contin-
gency as regards the relation of means to end. On the
assumption that instinct is inherited habit, the actions,
before becoming habitual, must have been performed
either intelligently or as a result of blind groping. In
either case, their adaptedness to the end in view was, at
the outset, accidental, as we have already seen. On the
assumption that instincts have arisen through natural
selection, chance tendencies toward movements of an
adaptive sort were perpetuated. Here, the complete
contingency is obvious, unless we assume some directing
influence determining the nature of the variations. I
shall return to this last point later.

Still lower than instincts, in the scale of organic beha-
vior, we have the various responses to stimuli which are
known as ‘‘tropisms’’ or ‘‘taxes.’’ Under this head are
included the locomotion of the organism to or from a
source of stimulation, or, in the case of a fixed organism,
the assumption of a definite position, or the arrangement
of its parts, with relation to the direction of the stimulus.

Here, again, we have two rival hypotheses, which are
not, it seems to me, wholly antagonistic. According to
one view, the organisms are ‘‘fatally’’ turned to or from
the source of light, heat, or the like by the unequal stimu-
lation of the opposite sides of the body. When the ap-
propriate orientation has been brought about, the two
sides of the organism are equally affected, and further
locomotion will be in line with the source of stimulation.

The other view lays stress on those cases in which or-
ganisms are not drawn directly towards or away from a
stimulus, but undergo random movements, having no
primary relation to it. In whatare regarded as the most
primitive cases the stimulus which results in a change

No. 626] ADAPTATION 199

of behavior is usually a noxious one, leading to a backing
out or turning aside. When forward movement is re-
sumed, it is a matter of chance whether the organism
remains in favorable surroundings or finds its way back
to the unfavorable ones. If the latter, the ‘‘avoiding
reaction’’ recurs, and the performance is repeated until
it leads to a more fortunate issue. The invariable
‘‘pull’’ or ‘‘push”’ of the tropism theory is not regarded
as the primary phenomenon, though an observer who
viewed only the end results of the process might easily
believe that they had been brought about by such a di-
recting influence.

Here, again, it is not my purpose to discuss the merits
of these rival hypotheses. It is possible, indeed, that
they should be regarded as complementary, rather than
antagonistic. Jennings admits that responses which
originally were performed according to the method of
‘‘trial and error” may, through the abbreviating influ-
ence of habit, come to be determined more directly by
the stimulus. But, however we may view the method of
origin of these responses to stimuli, it seems plain that
any adaptiveness that we meet with is contingent in the
sense in which I have already used the term. According
to the investigations of Jennings, the organism reaches
an optimum environment by chance, and remains there
because it is stimulated to change its course whenever it
begins to pass out of this environment. That unfavor-
able stimuli should provoke these changes of behavior
need not be attributed to any ‘‘primary purposefulness’’
in living matter, since we can be perfectly sure that any
organisms behaving differently would be speedily elimi-
nated.

Again, I take it that the chief advocate of the theory
of direct orientation would be the last to assume a prin-
ciple of primary adaptedness, and would admit that any
utility connected with these ‘‘tropisms’’ must have been,
in the first instance, a pure coincidence. In the case of
an organism, ‘‘irresistibly’’ drawn toward a favorable

200 THE AMERICAN NATURALIST [Vou. LIII

stimulus, as a growing plant toward the light, we might
seem to have an instance of such a directly purposive
action, 7. e., the determination of the means by the end.
But several things must here be taken into consideration.
(1) It not infrequently happens that organisms are
drawn in an equally irresistible manner toward a fatal
stimulus, e. g., the moth to the flame; (2) we can not feel
sure, in every case, that the attainment of the goal is not
the outcome of random movements, unperceived by the
observer; (3) even where the response is indubitably
adaptive, and as direct and unfailing as a simple reflex,
it may be the outcome of a mechanism developed through
natural selection, i. e., the survival of random variations
which were as frequently unadaptive as they were adap-
tive. The fact that some organisms still make suicidal
responses to less familiar stimuli favors this last view.

Next, we may consider the phenomena of metabolism,
growth and development. We group these things to-
gether, because they can hardly be considered separately.
Growth is the outcome of metabolism, and development
of metabolism and growth.

The phenomena revealed through studies of normal
physiology and embryology are obviously highly ‘‘pur-
posive,’’ in the sense that they have relation to the at-
tainment of an end, that end being the preservation of
the individual and the race. Nevertheless, they are be-
lieved by most biologists to be the outcome of a ‘‘mech-
anism’’ the functioning of which presents no greater
difficulties, apart from complexity, than the working of
a clock or a steam-engine.

When we come to consider the origin of this mech-
anism, we may mention three chief hypotheses, which
have been or still are held. (1) It may have been spe-
cially created by a super-mundane power in each indi-
vidual species of organism; (2) it may have gradually
developed out of simple beginnings by the ‘‘selection’’
or survival of random variations which were as likely to
be unadaptive as adaptive; or (3) it may have gradually

No. 626] ADAPTATION 201

developed out of simple beginnings through (a) direct
responses to environmental stimuli, or (b) the effects of
functioning upon the functioning parts themselves.

The first of these alternatives has been well-nigh dis-
carded by scientist and layman alike, and need not be
further considered here. I will point out in passing,
however, that certain elements of the ‘‘special creation’’
hypothesis have recently been put forward in the name
of science. Of this more anon.

The second alternative, that of natural selection, is ad-
mitted by most biologists to be one of the factors con-
cerned in the production of adaptive mechanisms, though
it is. doubtful whether any two thinkers would agree as
to the importance to be assigned to it. The essence of
this hypothesis is the contingency of the individual vari-
ations in relation to the need to be satisfied. If the vari-
ations are directed, in the sense of tending preponder-
atingly toward the satisfaction of this need, then our
explanation is shifted to a totally new basis. It is this
directive tendency, not natural selection, which is the
effective agency in evolution. The consequences which
would follow such an assumption will be discussed later.

The third of our alternative hypotheses has figured his-
torically as the chief rival of natural selection, though by
many (e. g., by Darwin himself), both principles were
accepted. One of the merits of the Lamarckian prin-
ciple, in the eyes of somé of its adherents,’ is its apparent
rejection of contingency or chance, a fatal weakness, so
they believe, in the natural selection theory. But a little
thought will show us that the Lamarckian principle, no
less than the Darwinian, is based upon chance, as regards
the relation between the need and the means to its ful-
filment.**

3 August Pauly, ‘‘Darwinismus und Lamarckismus,’’ 1905; R. H. Franeé,
‘‘ Der heutige Stand der Darwin’schen Fragen,’’ 1907

8° It is true that both of the writers cited in the preceding footnote clearly
recognize this accidental character of adaptive responses in their inception,
though failing to realize the significance of this fact for biological philos-
ophy. (See my review of Pauly, in Journal of Philosophy, Psychology and
Scientific Methods, August 27, 1908.)

202 THE AMERICAN NATURALIST [Vor. LII]

Let us consider (a) the case of modification through
direct environmental stimuli. There is much vague talk
about the ‘‘environmental mould,’’ in which the ‘‘plas-
tic’? organism is supposed to be ‘‘cast’’; but those who
have given much study to the subject recognize that modi-
fications produced by the environment are in the nature
of reactions to stimuli. In many cases, these reactions
are plainly adaptive, in the sense of furthering the life
or comfort of the individual or the race, as when a cal-
losity is developed in consequence of continued friction,
or an antitoxin is generated to combat a bacterial poison.
The fact, however, that there are varying degrees in the
adaptiveness of these responses, and indeed that many
of them appear to be wholly unadaptive, suggests the
probability that the truly adaptive ones, when at all con-
stant, have resulted from the selection of ‘‘accidental’’
variations. This can, of course, be true only of re-
sponses to environmental stimuli which have presented
themselves frequently in the history of the race. Cases
in which the organism has responded adaptively to stim-
uli quite new to racial experience are not, however, en-
tirely unknown. These will be discussed in a later sec-
tion. It may be said in passing, however, that the only
conceivable scientific explanation of such cases involves
the principle of ‘‘trial and error,’’ which, of course, is
based upon complete contingency as regards the relation
of means and end.

Let us pass to (b) the effects of functioning upon the
functioning parts themselves. Itis held by the Lamarck-
ians that organs or parts grow or diminish through use
and disuse, and that the perfected mechanisms which
now arouse our admiration and wonder are the outcome
of past functional activity. Many upholders of this view
introduce the idea of a conscious struggle toward a de-
sired end. The perfecting of the parts they regard as
a voluntary process. Thus is ‘‘blind chance’’ cast out
as a factor in evolution. But such reasoning rests on
insufficient analysis. Granting the part played by vol-
untary action (e. g., exercise or practice) in the post-

No. 626] ADAPTATION 203

natal development of many higher organisms, we need
only refer the reader to what we have already said about
the ‘‘contingent’’ character of even intelligent action.
But it seems likely that the claims of the psycho-vitalists
(e. g., Pauly and Francé) are largely fantastic, and that
voluntary struggle toward an end has not played the im-
portant rôle in organogenesis which they imagine that
ithas. The greater part of the functioning of the organ-
ism probably consists in blind responses to external or
internal stimuli—blind in the sense of having no con-
scious end in view. Thus regarded, they are in no way
different from the responses already considered under
(a), save that we there dealt with the effects of external
stimuli alone.

Accordingly, we may repeat here that so far as these
functional responses—and the organs they perfect—are
adaptive, their adaptiveness must have arisen, in the first
instance, by the selection of contingent variations. Un-
der this head are to be included (1) the preservation of
those individuals which chanced to make appropriate
responses (natural selection) ; and (2) the making habit-
ual on successive generations of individuals of responses
which chanced to fulfil a given need when first expe-
rienced (Lamarckism). The only other alternative would
seem to be some sort of inscrutable foreknowledge on the
part of the organism of every need to be experienced, and
of the way in which this need could be satisfied. Such a
conception would obviously carry us beyond the field of
scientific explanation, but I shall none the less consider
it in its proper place.

It is not my purpose here to discuss the arguments for
or against either the Darwinian or the Lamarckian prin-
ciple. It is my object merely to point out that both
theories rest on the selection, in one way or another, of
variations which were originally contingent or acci-
dental, in the sense of not being directly determined by
the need to be fulfilled. And, indeed, this is true of all
of the other rival or subsidiary hypotheses of evolution,

204 THE AMERICAN NATURALIST [Von. LIII

so far as they may be regarded as scientific theories
at all.

The theory of mutation, in its original form, postulated
large and abrupt variations as the material for selection,
a modification which does not affect the principle essen-
tially. In its later form, it merely insists that these
variations must be of the discontinuous or Mendelian
type, assuming that all other variations are non-inheri-
table. Those who maintain the importance of isolation
in evolution can not, of course, regard this as a vera
causa of adaptive change. The actual changes must be
either ‘‘spontaneous’’ variations or mutations or else
modifications due to environment. Thus, we must resort
finally to either the Lamarckian or the Darwinian prin-
ciple to account for such of them as prove to be useful.
‘‘Orthogenesis,’’ so far as it is not a vague appeal to a
‘‘perfecting principle,’’ ‘‘élan vital’’ or the like, is a mere
assertion that variations may accumulate in a given
direction independently of selection. Wherever the vari-
ations are sufficiently adaptive, however, we are not justi-
fied in excluding selection. When non-adaptive, such a
process presents no greater difficulty in principle than the
continuous growth of a crystal or the continuous deepen-
ing of a canyon by erosion.

Most of us are prepared to admit that much in the
organic world is non-adaptive. We may even grant that
a large proportion of the diagnostic characters of species
and genera belong to this category. Such characters,
while they may baffle the investigator, are in general not
such as would have suggested the operation of a super-
natural factor in evolution. In this paper we are con-
cerned with the problem of organic adaptation, and shall
leave aside the origin of characters which are useless to
the organism.

In the foregoing analysis, I have regarded adaptive
response, whether of structure or function, as being in-
variably a secondary phenomenon. The connection be-
tween the need of the organism and the means adequate
to satisfy this is believed to have always been, at the out-

No. 626] ADAPTATION t 205

set, an ‘‘accidental’’ one. In those cases where the cor-
rect response appears to ensue unhesitatingly, we have
had to suppose either (1) that the observer has overlooked
‘‘trial and error’’ stages preceding the response in ques-
tion, or (2) that the response is the outcome of an inher-
ited mechanism, based upon racial experience, and there-
fore ultimately upon some form of selection.

III. Virarism‘4

Let us now consider the claims of a school of thinkers
who argue for the existence of a primary purposefulness
in living things, and who deny that any conceivable mech-
anism can account for certain of the phenomena ob-
served. As the most conspicuous representative of this
school we naturally turn to Hans Driesch, who has made
a more determined attempt than any other vitalist to
reduce his beliefs to a unified system of philosophy.

Driesch’s three ‘‘proofs’’ of vitalism may be summar-

„ized as follows:

1. In the earlier development of some organisms,
rather low in the scale of life, any part of the embryo,
provided that it be of a sufficient size, will, if artificially
detached, produce the entire organism. This he regards
as conclusive disproof of the supposition that the spa-
tially arranged diversities of the adult organism depend
for their origin upon diversities of a spatial sort in the
embryo. Such a spatial prearrangement of the parts as
is postulated by the Weismannian ‘‘germ plasm” theory,
and other preformationist hypotheses, he assumes to be
essential to any mechanical theory whatever.

But, Driesch claims, the spatial diversities of the adult
organism must depend upon preexisting diversities of
some sort, therefore he invokes a non-spatial agent, ‘‘en-
telechy,’’ to account for them. Now ‘‘entelechy’’ must
be a manifoldness, since it is conjured up to explain
other manifoldness, but this manifoldness is intensive,

4In this section, I have made free use of a review of Driesch’s ‘‘Science
and Philosophy of the Organisms,’’ which I wrote some years ago (Journal
of Philosophy, Psychology and Scientific Methods, June 9, 1910). I have
not thought it necessary, however, to indicate the extent of these quotations.

206 THE AMERICAN NATURALIST [Vou. LII

not extensive. As an illustration of an ‘‘intensive mani-
foldness’’ he instances one of our own states of con-
sciousness, in which many elements are presented simul-
taneously, though not spatially separated from one an-
other. But entelechy is not to be identified with mind.
It is an unknown something which stands in the same
relation to our mental me as it does to other organic
phenomena.

2. Driesch’s next ‘‘proof’’ of vitalism is somewhat
similar to the first, though it rests upon the facts of nor-
mal life history, instead of upon artificial disturbances
of this. The primitive germ cells, each of which, accord-
ing to the hypothesis he combats, should contain the
‘‘machine’’ or spatial prearrangement of parts neces-
sary for the development of an entire organism, undergo
in the gonads an extensive series of divisions, leading to
the formation of the mature ova and spermatozoa.
‘‘Can you imagine,’’ he asks, ‘‘a very complicated ma-
chine, differing in the three dimensions of space, to be:
divided hundreds of times and in spite of that to remain
always the same whole? ’”

3. The last ‘‘proof’’ of vitalism is based upon an anal-
ysis of animal behavior. Driesch makes much of the
fact that an action of a higher animal, particularly of an
intelligent one, is something more than the sum of many
simpler elements, each depending upon an element in
the complex of stimuli to which the organism responds in
a given case. The response of the organism is a unified
whole, corresponding to a total situation in the outer
world. A slight change in this complex of physical stim-
uli, provided that it has significance for the organism,
may result in a totally different kind of response. On
the other hand, an entirely different set of physical ele-
ments—having, however, the same meaning for the or-
ganism—may call forth precisely the same response. In
other words, there is no functionality (in the mathemat-
ical sense) between the response and the stimulus.

This line of argument, different as it may seem, rests

5‘*Seience and Philosophy of the Organism,’’ Vol. I, p. 225.

No. 626] ADAPTATION 207

_ upon the same fundamental assumption as the two pre-
ceding ones, namely, that a truly mechanical theory must
find in the cause as many separate elements as we ob-
serve in the effect. The structural diversities of the
adult organism must rest upon corresponding structural
diversities, present from the beginning in the germ. The
functional diversities, constituting a complex act of be-
havior must rest upon corresponding functional diver-
sities in the stimuli which make up the total effective
situation. If no such correspondence can be shown, we
must invoke some principle of a totally different nature
from those which we employ as explanations in ~~ inor-
ganic world.

Now, such a conclusion as this seems to rest upon an
insufficient consideration of what really happens in the
inorganic world. In a sense, the solar system was pres-
ent potentially in the original homogeneous nebula, while
the various continents and oceans, mountains, lakes and
rivers of the world we live in were all present potentially
in the molten globe which in some way detached itself
from the parent mass. But there was certainly no ‘‘pre-
formation’’ of these final products of cosmic evolution.
The diversity which was introduced was totally new. In
the language of biology the world’s development was
strictly ‘‘epigenetic.’’ And yet the process was none the
less mechanical, as every vitalist will allow. Why then
does Driesch insist that a mechanism adequate to ac-
count for an animal’s ontogeny must present a part-for-
part correspondence with the adult organism? For it is
only a mechanism, as thus conceived, that is disposed of
by his ‘‘proofs’’ of vitalism. His experiments compel
him to dismiss the notion of a spatial prearrangement of
parts. Therefore, he jumps to the conclusion that there
must be a non-spatial prearrangement of parts—an ‘‘in-
tensive manifoldness.’’ But why should there be any
prearrangement of parts at all? Is it not a fallacious
philosophy which insists on such an exact numerical cor-

208 THE AMERICAN NATURALIST [Vou. LIII

respondence between the elements of the cause and the
elements of the effect ?°

Despite these logical difficulties, Driesch’s third
‘‘nroof’’ of vitalism contains such an unmistakable ele-
ment of plausibility that some further consideration may
profitably be given to it here. His contention is summed
up in the phrase ‘‘individuality of correspondence’? be-
tween stimulus and reaction. ‘‘It is not the single con-
stituents of the stimulus,’”’ he says, ‘‘on which the single
constituents of the effect depend, but one whole depends
on the other whole, both ‘wholes’ being conceivable in a
logical sense exclusively” (II, 81). Why is it that we
react to objects rather than to sensuous images? ‘‘The
dog, ‘this dog,’ ‘my dog,’’’ to quote Driesch, ‘‘is ‘the
same’ stimulus, seen from any side or at any angle what-
ever: it always is recognized as ‘the same,’ though the
actual retinal image differs in every case” (II, 73). Ex-
perience and association, he thinks, afford an insufficient
basis of explanation here. There must be something ca-
pable of resolving past experience into its elements and
making wholly new combinations of them.

Driesch challenges his opponents even to conceive of
a machine that could accomplish results such as these.
This introduction of the word ‘‘machine’’ would seem to
prejudice the case in his favor at once. But is he not
really challenging us to imagine how phenomena that re-
quire sense organs and a nervous system for their per-
formance could be performed by some other type of
mechanism which is simpler and more fully understood
by us. Confessedly we can not do so. Looking at the
subject in an unbiased way, it would seem that the nerv-
ous system had the appearance of a finely wrought mech-
anism to a higher degree than any other portion of the
body. It is truly one of almost infinite complexity, and
one that is largely inaccessible to experimental observa-
tion. But certain significant facts have been demon-

6 Spaulding (Philosophical Review, July, 1909) and Jennings (Johns
Hopkins University Circular, No. 10, 1914) have already called attention to
the fallacy of this aspect of Driesch’s argument.

No. 626] ADAPTATION 209

strated none the less. Sherrington has described in
some degree the mechanism of inhibition, and has ascer-
tained some of the factors which determine which of two
simultaneous stimuli shall prove effective in a given re-
flex. Do not such data at least help us to conceive the
possibility of a nervous system whose activities may be
understood without the aid of an entelechy to make its
decisions for it?

The emancipation of the organism from the controlling
influence of immediate stimuli is admitted to be one of
the salient features in animal evolution. Now, in order
that present activities may be directed with reference to
future results, the stimuli must become more and more
symbolic, i. e., they must acquire a ‘‘meaning.’’ That
one thing may ‘‘stand for” something else, and call up
the responses proper to that something else may readily
be understood in terms of association. At least there
would seem to be no desperate need for invoking ‘‘en-
telechy’’ at this point. If this be granted, why should
we expect any correspondence between the sensuous ele-
ments of the stimulus and the elements of the response?
The effects of a given ‘‘individualized stimulus”’ are de-
pendent rather upon the aggregate of associative proc-
esses which this stimulus calls up. And this aggregate
is altogether an empirical one, not a logical one as
Driesch supposes. The connections that bind it together
may be quite arbitrary and accidental. It is partly the
product of individual experience, partly of racial expe-
rience—this last on any theory of inheritance. That
Several widely different stimuli, having the same mean-
ing (7. e., having certain important associations in com-
mon), can bring about essentially the same response
would seem, on the face of it, no more difficult to under-
stand ‘‘mechanically’’ than that several very differently
Shaped keys can open the same lock.

The weakness of Driesch’s ‘‘third proof of vitalism’’
would seem, therefore, to be twofold. (1) He appears
to believe that an explanation, in order to be mechanical,

210 THE AMERICAN NATURALIST [Von LIII

must find a definite correspondence between separate fac-
tors of the cause and separate factors of the effect,’ and
(2) he appears to believe that in any mechanical expla-
nation of action the character of the response must be
determined by the immediate sensuous stimuli them-
selves, without regard to the representative (associa-
tional) character of these stimuli.

Driesch, like other vitalists, lays great stress upon
“adaptive”? or ‘‘regulative’’ phenomena, though he
makes no claim that these necessarily demonstrate the
truth of vitalism. Indeed, it will be noted that the three
foregoing ‘‘proofs’’ rest on quite other grounds. We
may safely say, however, that for most biologists the
great stumbling-block to a consistent mechanical expla-
nation has been this central fact of organic ‘‘purposeful-
ness.” In pre-Darwinian days the whole subject was a
mystery, which science cheerfully handed over to theol-
ogy for solution. Later, we grew accustomed to the idea
that much which seemed purposeful in nature was the
outcome of ‘‘chance.’’ But for many there was always
a considerable residuum which defied solution. For
there certainly seem to be cases of adaptive response to
wholly new situations, that can not be accounted for on
the basis of an evolved mechanism. And furthermore, it
is now obvious that no single theory of evolution yet
proposed, nor, indeed, all of them combined, can ade-
quately account for much that has come to pass.

In the face of these perplexities, it is but natural that
many have taken refuge once more in various intangible
forces and principles, almost wholly devoid of positive
attributes, and agreeing only in their alleged competence

7It must be admitted that explanations of this type have been put for-

ard by avowed mechanists. Thus Loeb (‘‘ Mechanistic Conception of
Life,’’ p. 80), in discussing the present writer’s experiments upon the color
changes of fiatfishes, concludes that there is an brste reproduction on the
skin, through the brain, 3 er retinal images ackground. I think

that a careful reading of wn discussio Kohi experiments sufficiently
disposes of this sited jesse of Eaperimenia Zoology, May, 1911).
The yet more extensive experiments of Mast (Bulletin of the Bureau of

Fisheries, Vol. XXXIV, are likewise conclusive against this vi

No. 626] ADAPTATION 211

to ‘‘explain’’ otherwise inexplicable facts. Of oes
Driesch’s ‘‘entelechy’’ and Bergson’ s “élan vital” a
but types.

Much less discordant with our scientific habits of
thought are the utterances of some of the so-called
‘*psycho-vitalists,’’ to whom allusion has already been
made. These writers do not have recourse to meta-
physical principles, wholly beyond the realm of expe-
rience. They invoke the familiar facts of conscious pur-
pose, intelligence and will. Organic happenings seem
purposive, they think, because they are purposive, in the
same sense that our own voluntary actions are purposive.
Such a view carries the realm of mental life far beyond
the bounds which we are wont to assign to it. Its logical
outcome is a thoroughgoing panpsychism, an outcome
which some of its advocates are quite ready to accept.

Now, it seems to the writer that a panpsychie view of
nature can be stated in such terms as not only to be
plausible, but to meet certain of our most fundamental
intellectual needs. But such a view is at best a philo-
sophical creed, not a scientific explanation, and should
never be offered as a substitute for the latter.

The introduction of will, purpose, etc., in the rôle of
scientific explanations may have one of two implications.
Either (1) it may be assumed that a given physical con-
figuration, plus these psychical concomitants, is able to
accomplish what would be impossible for the same phys-
ical configuration minus these psychical concomitants
- (interactionism); or (2) it may be assumed that only
that type of physical configuration which is invariably
bound up with certain psychical factors is competent to
call forth the result in question (parallelism). Accord-
ing to the second point of view, the question whether the
same result would have ensued without the agency of
purpose or will is an absurdity. If purpose and will had
been lacking, the physical antecedents would of necessity
have also been different.

It is needless to say that both of the foregoing posi-
tions have been upheld by philosophers. Itis my wish

2iZ THE AMERICAN NATURALIST (Vou. LIII

to point out, however, that on neither assumption does
the introduction of conscious purpose supply a missing
link in our explanation of the ‘‘teleological’’ in nature.
Whether or not we admit the efficacy of mental states, in-
dependently of their physical concomitants, we have
already seen that conscios purpose must proceed on the
basis of experimentation. It must have learned through
trial that a given means will lead to the attainment of a
given end. The existence of any primary foreknowledge
of the relation of means to end is contradicted in our own
every-day experience.

t may be useful to introduce a description by a psy-
chologist® of what actually occurs when we are trying to
solve a problem:

ur only command over it is by the effort we make to keep the pain-
ful unfilled gap in consciousness. ... Two circumstances are impor-
tant to notice: the first is, that volition has no power of calling up
images, but only of rejecting and selecting from those offered by spon-
taneous redintegration [= association]. But the rapidity with which
this selection is made, owing to the familiarity of the ways in which
spontaneous redintegration runs, gives the process of reasoning the
appearance of evoking images that are foreseen to be conformable to
the purpose. There is no seeing them before they are offered; there
is no summoning them before they are seen. e other circumstance
is, that every kind of reasoning is nothing, i in its simplest form, but
attention.

It is, therefore, a false theory of our own purposeful
actions that is projected backward into organic nature
by the psycho-vitalists. The existence of instinctive
acts, which fit means to ends, prior to experience, in no
way invalidates what I have said. For these may be
assumed to be based, in some way, on past racial expe-
rience. And, in any case, so far as an action is instine-
tive, it can not be consciously purposive. Assuming that
instinctive actions are performed consciously, at all,
which some would perhaps deny, it is not likely that any-
thing beyond the next succeeding step in the series is at
any moment present to consciousness. The biological
meaning of the entire performance (say the building o

8 Hodgson, quoted by James (‘‘ Principles of Psychology,’’ Vol. I, p. 589).

No. 626] ADAPTATION 213

a nest) can not be understood by the organism. Each
step is desired and willed on its own account alone. The
end takes care of itself, by virtue of a preestablished
mechanism. 3

Thus purpose, in the psychological sense of the word,
can not be predicated of a complex instinctive act, even
though the individual steps be consciously performed.
Least of all can it be predicated of a process of organic
regulation or reparation, the object of which can never
be consciously in view. It was doubtless in part consid-
erations like these which led Driesch to deny the mental
nature of ‘‘entelechy’’ altogether and to remove it to a
transcendental sphere in which it was no longer subject
to the exacting demands of experienced reality. Indeed,
he tells us that ‘‘there must be a something in them
{morphogenetic, adaptive and instinctive entelechies]
that has an analogy not to knowing and willing in general
. . . but to the willing of specific unexperienced realities,
and to knowing the specific means of attaining them’’
(II, p.142). We think more favorably of Driesch’s good
sense when he admits that the position of his doctrine is
at this point ‘‘rather desperate.’’ Nor is Bergson’s
case a bit better when he naively attempts to clear up
certain of the most baffling phenomena of instinct by in-
voking the aid of ‘‘intuition’’ or ‘‘sympathy.’” The
psycho-vitalists introduce an agent which is to some
degree intelligible, even though it is inadequate. The
agents which Driesch and Bergson conjure up are neither
adequate nor intelligible.

In the writings of these and some other vitalists the
‘‘vital principle,” by whatever name called, is distinctly
credited with powers which we should ordinarily term
clairvoyant. Indeed, we are forced to conclude that it
must be able to ‘‘tap’’ sources of information which are
closed even to the highest finite intelligence. This, of
course, is mysticism pure and simple, though such a re-
proach admittedly does not constitute its refutation for
all minds. )

9**Creative Evolution’’ (trans.), pp. 173-175.

214 THE AMERICAN NATURALIST (Von. LIT

It may be interesting perhaps to consider where such.
assumptions would lead us. Suppose that we adopt the
absolutist idea of an Infinite Knower, having cognizance
of the future as well as the past, or rather including both
future and past in one eternal present. By getting into
connection with this, our entelechy could doubtless solve
any problem which confronted it. But how, on such an
assumption, could we account for the multitudinous mis-
adaptations which confront us? How should we explain
an instinct which led to the harboring of baneful para-
sites in an ant community or a regenerative process
which resulted in the formation of the wrong organ?
Perhaps these perplexing cases would be merged into the
general mystery of the origin of evil, and there, indeed,
may be where they belong.

But we are not compelled to accept an absolutist inter-
pretation of things. As scientists, we may find it more
easy to believe in the evolution of God, in a ‘‘ Dieu qui se
fait.’’” Well and good, but then the essence of this view
is the newness of everything that happens. No mind,
however infinite, could foresee the future, for the simple
reason that the future is not determined until it comes to
pass. Even our deity must learn by experience, and
‘‘entelechy’’ would have to do the same. In that case
neither would be of much service in attempting to ex-
plain organic purposefulness. Had we previously learned
to expect any great amount of consistency among the
various views of M. Bergson, it would have been a source
of surprise to us to find him coupling together this idea
of ‘‘creative evolution’’ with a transcendental ‘‘élan
vital,” which provides the organism with useful struc-
tures without the guidance of experience.

Such a departure as I have made from the field of
legitimate scientific discussion may shock those of my
readers who shy at anything suggestive of metaphysics
or theology. But we have been told with increasing
frequency of late that our accepted scientific methods

10 T believe that this expression is Bergson’s.

No. 626] ADAPTATION 215

have broken down in the face of vital phenomena, and
that the only path of escape was one which logically led
to mysticism. For this reason it seemed worth while to
inquire whether even this abandonment of our scientific
principles would lighten our difficulties.

Now, while we believe the solutions offered by the vi-
talists to be but pseudo-solutions, we must admit that
the issues they have raised are real ones. It is to the
great credit of this school, and of Driesch in particular,
that they have awakened some of us biologists from our
‘‘dogmatie slumber’’ and forced these problems upon
our attention. The problems are real ones in the prag-
matic sense of determining our attitude, both theoretical
and practical, toward biological investigation in general.
Most important of all, vitalism has unearthed a number
of highly interesting experimental data, which it chal-
lenges its opponents to explain. To this extent it may
lay claim to the rank of a ‘‘working hypothesis.”

Let us consider some of the points at issue between
vitalism and what I shall call ‘‘scientifie biology.” In
what follows, I have stated what I believe to be the typi-
cal attitude of each side, though it is likely that no two
persons would agree in every particular.

1. Scientific biology is strictly deterministic. It ad-
mits the possibility of only one result from a given set
of antecedents. Vitalism is indeterministic, holding that
from precisely the same antecedent situation more than
one result is possible. Driesch saves the principle of
‘‘univocal determination’’ by saying that in cases where
different results follow the same physical causes, there
must have been a difference in ‘‘entelechy.’? But John-
stone,'' a disciple of Driesch, throws over even this
formal adherence to scientific method, and asserts boldly
that there must be ‘‘uncaused differences’’ in the organic
world. He illustrates this belief from the variability
among the millions of eggs spawned by a single flounder.
The usual explanation, based upon differences in exter-

11 ‘‘ The Philosophy of Biology,’’ 1914.

216 THE AMERICAN NATURALIST [Vou. LII

nal conditions or on imperfections in the mechanisms of
cell-division he holds to be inadequate. Now, in his be-
lief, it is these ‘‘spontaneous’’ variations (using the
former word literally) that furnish the raw material for
evolution.

Jennings!” sees in this postulate of indeterminism the
fundamental fallacy of vitalism. It certainly is the fea-
ture that would most seriously affect us as investigators.
For whether variations are regarded as uncaused or as
caused by an agent beyond the ken of scientific investiga-
tion matters little. Any attempt to account for them by
experimental or observational means must be futile.

2. Scientific biology endeavors to explain organic phe-
nomena on the basis of antecedent physical conditions,
though admitting that our knowledge of cause and effect
is in the last resort empirical, to the extent that much
which happens could not have been predicted in advance.
Vitalism explains organic phenomena—or a certain part
of them—on the basis of ends to be realized, and gives to
these ends a determining influence in providing the means
to their realization. Since the a tergo ‘‘push’’ of phys-
ical causation would only by rare chance be directed in
harmony with these ends, vitalism introduces a non-
physical agent to guide or control the former. Driesch
goes to great lengths to explain how ‘‘entelechy’’ can
play this rôle without coming into conflict with the law
of the conservation of energy.

In a certain sense the existence of such ‘‘ends’’ must
be admitted by all biologists. Attainment of the typical
form, self-preservation, racial preservation, ete., are
‘‘ends’’ in the sense that organic processes in general
are observed to trend in those directions. Furthermore,
disturbances of this normal trend often seem to be cor-
rected automatically. Phenomena strictly analogous in
this respect can, of course, be instanced from the inor-
ganic world. All we have to do is to designate the ob-
served goal of such a process as the ‘‘end’’ and the

12 Science, June 16, 1911; October 4, 1912.

No. 626] ADAPTATION 217

causally determined steps become the means to its reali-
zation. The difference between such a physical process
and a vital one, as conceived by Driesch, is that in the
latter a given sequence of events may or may not come to
pass, depending on the whim of ‘‘entelechy.’’ The issue
here, then, is practically the same as that first raised,
namely, that of determinism versus indeterminism.

3. Scientific biology declares that vital phenomena are
chemico-physical, in the sense that they are the inevi-
table outcome of the particular material aggregations
which we term organisms.!* It grants that in these
manifold chemical syntheses entirely new properties
have emerged, though insisting that the same may be
said of any union of elements whatever. Vitalism de-
nies that any possible configuration of material parti-
cles, without the aid of an immaterial principle, can
account for the phenomena observed. It is for this rea-
son that ‘‘vitalism’’ is commonly set in opposition to
‘‘mechanism.’’ Driesch’s three ‘‘proofs’’ of vitalism
are concerned with this last aspect of the theory. We
have seen that all three are based on the assumption that
we must find a diversity in the cause, corresponding to
each diversity in the effect. And it has been pointed out
that this does not hold true even of admittedly physico-
chemical systems.

Now, I do not claim that the bare word ‘‘mechanism,’’
however hallowed by scientific usage, has any greater
explanatory value than ‘‘entelechy.’’ Indeed, I do not
see why we should be called on to furnish a mechanical
explanation, sensu stricto, of biological phenomena at all.
Not all natural science is mechanics; some of it is chem-
istry. And I believe it is equally true that still another
part is biology, a science quite distinct from either. But
I think we can claim the possibility of a scientific explana-
tion in the sense indicated by the foregoing antitheses,
and it is with this in mind that I have grappled with the
problem of organic ‘‘purposefulness.’’

(To be concluded)

13 It is not, however, necessarily ‘‘ materialistic’ in a metaphysical sense.

GIGANTISM IN NICOTIANA TABACUM AND ITS
ALTERNATIVE INHERITANCE

H. A. ALLARD

TOBACCO INVESTIGATIONS, BUREAU oF PLANT INDUSTRY,
Wasuineton, D. C.

INTRODUCTION

WITHIN recent years observers working with different
varieties of Nicotiana tabacum grown commercially in
the United States and elsewhere have recorded the sud-
den appearance of occasional giant plants of abnormally
high leaf number. Except in height and number of
leaves, which may be increased several times above the
usual number, these giant plants in general appearance
do not depart widely from the varietal type from which
they took their origin. The great increase in number of
leaves, together with a greatly elongated main stem, is
accompanied by a period of vegetative vigor of such long
duration that blossoming does not normally take place
when the plants are growing in the field. In order to ob-
tain seed from such plants, the usual practise has been
to transplant the roots and stub, or even the plants en-
tire, to the greenhouse in the fall, where vegetative vigor
is resumed with the final production of normal blossoms
and seed during the winter. Plants of this habit of
growth have been recorded in the Sumatra, Maryland.
Cuban and Connecticut Havana types of tobacco.

OccURRENCE OF GIGANTISM IN DIFFERENT VARIETIES

The first published record of gigantism in tobacco ap-
pears to have been made in 1905 by Hunger (1905), work-
ing with tobacco in Sumatra in connection with an inves-
tigation of the mosaic disease.

Garner (1912) mentioned a Maryland Mammoth type,
the origin of which was associated with a cross between
two common varieties of Maryland tobacco.

Hayes and Beinhart (1914) reported the occurrence of

; 218

No. 626] GIGANTISM IN NICOTIANA TABACUM 219

giant plants in the Cuban shade tobacco and the Connec-
ticut Havana type in Connecticut.

In addition to Hunger’s observation previously men-
tioned, Honing (1914) brought out other interesting facts
concerning the oceurrence and behavior of giant plants in
Sumatra (Deli) and Java.

Hayes (1915) further discussed the occurrence of giant
plants in the Cuban and Connecticut Havana types of
tobacco grown in New England.

Hunger, in the paper referred to, states that the largest
giant plant observed by him developed 123 leaves and
. reached a height of nearly five meters. These plants were
entirely sterile, or, if blooming took place, the number of
blossoms was greatly reduced. Honing states that the
behavior of these giant Sumatra plants with respect to
the transmission of their peculiarities is variable. In one
instance he observed that a line of these plants finally
disappeared entirely. With respect to number of leaves,
Honing’s studies of the Deli tobacco indicates that sev-
eral more or less distinct types exist. Even though line
selections of these have been grown under bag for several
generations, plants possessing high leaf number have oc-
casionally appeared. Mammoth plants have also ap-
peared in the Sumatra variety grown in the United States
from seed obtained from Sumatra. In 1912 two plants
of this type appeared in a plot of about 100 plants grown
at Arlington, Va. These plants appeared in the second
year’s planting from seed obtained from Sumatra. One
of these, when removed to the greenhouse, had reached a
height of eleven feet and had produced about 100 leaves,
with no indication of blooming. It was not possible to
determine to what extent these plants transmitted their
characteristics to their progeny since both died after
being cut back and removed to the greenhouse.

In 1906 and 1907 giant or mammoth plants were ob-
tained in Maryland tobacco, as mentioned above. The

1A discussion of the commercial value of these types of Maryland tobacco
will be found in Bulletin 188, of the Maryland Agricultural Experiment Sta-
tion, entitled, ‘‘Types and Varieties of Maryland Tobacco,’’ by W. W.
Garner and D. E. Brown, 1914, pp. 135-152.

220 THE AMERICAN NATURALIST [Vou. LIII

type known as the Broadleaf Mammoth was first observed
in 1906 in a selection line of Maryland Broadleaf begun
in 1904. Of 100 plants grown in 1906, five were typical
mammoth plants producing many leaves and showing no
tendency to bloom at the end of the season. Subsequent
generations of these plants were grown successively in
1907, 1908 and 1909, and all reproduced the characteristic
habits of growth of the original parent isolated in 1906.
This mammoth type, as the name indicates, differed ma-
terially in shape of leaf from the better known Narrow-
leaf Mammoth.

The so-called Narrowleaf Mammoth appeared in 1907
in second generation plants of a cross made in 1905 be-
tween a Broadleaf type and a Narrowleaf type of Mary-
land tobacco. From a single mammoth plant found in
1907, 157 plants were grown in 1908, all of which were
mammoth plants. Two of these plants which were al-
lowed to grow until frost without topping had produced
109 and 111 leaves, respectively, with no indication of
blooming. The Narrowleaf Mammoth has been propa-
gated from seed and grown on a commercial scale in
Maryland up to the present time, and under normal field
conditions still retains its characteristics of high leaf num-
ber and the non-blooming habit.

A third mammoth type appeared in 1907 in second gen-
eration plants of a cross made in 1905 between Maryland
Broadleaf and the White Burley variety of Kentucky. In
a crop of 30,000 to 40,000 plants but one mammoth plant
was found. Unfortunately, this plant was harvested in-
advertently by laborers and lost.

From the previous discussion it is evident that gigant-
ism has occurred rather widely in the varieties of Nico-
tiana tabacum. It would appear from Honing’s work
that Mammoth Sumatra plants are not constant in their
inheritance and that intermediate forms exist. The ac-
cumulated experience of various observers working with
all Mammoth types which have appeared in the United
States, however, has shown a constant inheritance of

- No. 626] GIGANTISM IN NICOTIANA TABACUM 221

Mammoth characteristics from generation to generation.
Intermediate forms have not been observed.

BEHAVIOR OF GIGANTISM IN CROSSES

Since Mammoth forms are now grown commercially in
the United States and promise to become valuable new
varieties, it has been considered desirable to determine
the possibility of combining the Mammoth character of
indeterminate growth or gigantism with other characters
of commercial value by crossing Mammoth types with
ordinary varieties.

The Maryland Narrowleaf Mammoth has been crossed
with a number of pure lines of the more distinct varieties
of Nicotiana tabacum, including White Burley, Yellow
Pryor, Little Oronoco, Connecticut Broadleaf, and the
very distinct variety known as N. Chinensis (S. P. I., No.
42,355). In all these crosses the Mammoth characteristic
behaves as a unit character and is recessive to normal
size and normal blossoming habit of the ordinary
varieties.

A Maryland Mammoth and a Burley Mammoth, secured
as the result of the cross Maryland Mammoth 2? X White
Burley g, have also been crossed with the distinct species,
N. sylvestris and N. glutinosa. In these crosses the F,
plants invariably have blossomed normally as where
crosses were made with varieties of N. tabacum.

Under normal field conditions, first generation plants
of all Mammoth crosses have blossomed in practically
the same period required by the ordinary varieties of
N. tabacum. The plants, however, are usually somewhat
taller and, on an average, produce a somewhat higher leaf
number than the ordinary varieties, showing that the F,
plants are more or less intermediate between the normal
and the Mammoth parents. This relation of leaf number
is shown in Table I.

In crosses between Little Dutch and Maryland Mam-
moth, the F, plants were also somewhat larger and pro-
duced more leaves than the Little Dutch parent. F,

222 THE AMERICAN NATURALIST [Vot LIII

TABLE I
COMPARISON OF NUMBER OF LEAVES OF F, PLANTS OF CROSSES BETWEEN
MARYLAND MAMMOTH AND NORMAL ‘VARIETIES

| Leaf Number Classes

Variety | |

23 | 25 | 27 | 29 | 31 | 33 | 35 | 37 | 39| 41
Vellow Pryor (ai 3. OLUR O eit ay pre s |
Md. Marong 9 X Yellow Pryor 7... 1 | Lei V1 61-71-2313
Ditele-Oronoce . os eo ee A oS et 1} 1)}2)3) 3]
Md. Ma amak g XiLittle Orol AP wale Hae by Geb p Bid jaded pe
WTO TR i al alae aes) Ne 50° Sp ac 8 8S 114 | | |
Md. Mammoth 9 X White Burley 7... | wae | 1/6/10) 5}1
EEEN nena Serene ` EEE NERA falin REE BE eae ae Se AR aor seas

plants of the cross Maryland Mammoth 2 X N. Chinen-
sis & (S. P. I., 42,355) were grown in 1918, and records
of dates of blooming were made for comparison with the
dates of blooming of the parent N. Chinensis, which is an
unusually small and early maturing variety of N. tabacum.
From the following table it is evident that the parent N.
Chinensis blossomed somewhat earlier than the F, plants
of the cross with Maryland Mammoth:

TABLE II
eae or Days ELAPSING FROM sade To DATE oF First BLOOM
F F, PLANTS OF Cross MARYLAND MAMMOT X N. CHINENSIS J

AND PLANTS OF THE PARENT AS or N. CHINENSIS

Classes

45 | a7| 49 | 51 | 53 | 55 | 57 | 59 | 61 | 63

N. Chinesa POL a L a [88 11/2 0/2
Md. Ma. Mammoth Q Poise) Chinensis g . | | 1 2 | 1

In the cross Maryland Mammoth X White Burley,
Mammoth Burley types have consistently appeared in the
F, progenies, and have since remained true to Mammoth
character. These have been crossed with a number of
different types and varieties of N. tabacum. During the
summer of 1918 considerable data were secured at Arling-
ton, Va., showing the segregation of plants of Mammoth
dhetete i in the F, of many crosses.

Let us first consider the behavior of different Matnainokh
types when intercrossed. In these lines the Maryland

6

10
1 14

1

5
11

No. 626] GIGANTISM IN NICOTIANA TABACUM 223

Mammoth (narrowleaf type) has been crossed with Stew-
art Cuban (a giant type previously mentioned as origina-
ting in Connecticut in Cuban shade-grown tobacco), and
also with a Mammoth Burley type, which was secured in
the F, generation of the cross Maryland Mammoth X
White Burley. In the cross Maryland Mammoth X Stew-
art Cuban, many plants of the F, generation were grown,
all of which were of Mammoth habit of growth. Selec-
tions of these F, plants were grown and bred true to the
Mammoth habit.

In the cross Maryland Mammoth 2 X Burley Mam-
moth gf many F, plants were grown at Arlington, Va., in
1918. Of a total of 558 individuals, all were of Mammoth
habit and of this number twenty-one were yellowish
green like the normal White Burley variety, and 237 were
full green in color like the Maryland Mammoth parent.

In a study of the reappearance of Mammoth types in
the F, generation of crosses involving Mammoth and nor-
mal forms, several different combinations have been
made. In one group both parents were of Burley type.
In the second group one of the parents was normal green
and the other of Burley type. In the third group both
parents were green.

In the first group, involving Burley color in both pa-
rents, one of the parents was the Burley Mammoth se-
cured in the F, generation of the cross Maryland Mam-
moth X White Burley. From the cross Mammoth
Burley 2? X ordinary White Burley 4, 638 F, plants were
grown, of which 158 were Mammoth. This is a very close
approximation to the theoretical Mendelian ratio 638/4 =
159.2, which should obtain in a cross involving two sim-
ple contrasted Mendelian characters.

From the cross White Burley type of 30A,? 2 X Burley
Mammoth 2, 348 F, plants were obtained, of which eighty
were of Mammoth habit of growth. This figure closely
approximates the theoretical Mendelian ratio 348/4 = 87.

2 The type designated as White Burley type of 30A is a tall, vigorous

Burley type originally obtained from the eross Connecticut Broadleaf x
White Burley.

224 THE AMERICAN NATURALIST [Vou. LIII

Of the total number of Mammoth plants, i. e., 986, appear-
ing in the F, of these crosses, only two were Green Mam-
moth, the rest being typically of Burley character.
Whether these two exceptions represent mixtures or re-
versions can not be stated.

In the second group, one of the parents involved in the
original cross was Green, the other being of Burley
character.

From the cross Connecticut Broadleaf 2 X Burley
Mammoth J, 305 F, plants were grown, of which sixty-
nine were of Mammoth habit. This approximates the
theoretical ratio 305/4 = 76.2. From the cross Maryland
Mammoth 2 X White Burley type of 30A g, 152 F, plants
were grown, of which forty were of Mammoth habit. This
figure is very close to the theoretical ratio 152/4 — 38.
Of the total number of Mammoth plants, i. e., 457, which
appeared in the two crosses Connecticut Broadleaf 2 X
Burley Mammoth ¢ and Maryland Mammoth 9 x White
Burley type 30A 3, only two were of Burley color, the rest
being green.

We will now consider the third group, which involves
normal green color in both parents.

From the cross Connecticut Broadleaf 2 X Maryland
Mammoth ¢, 175 F, plants were grown, of which thirty-
nine were of Mammoth habit.

From the cross Maryland Mammoth ? X Yellow
Pryor ¢, eighty-three F, plants were grown, of which
twenty-five were Mammoth.

From the cross Little Dutch 2 X Maryland Mammoth J,
119 F, plants were grown, of which twenty-eight were
Mammoth. A total of 377 plants were grown in these
crosses, of which ninety-two were Mammoth plants. This
is a very close approximation to the expected ratio
377 /4 = 94.2.

Considering all the crosses in the three groups involv-
ing the Mammoth character in one of the parents a total
of 1,820 F, plants were grown, of which 439 were of
Mammoth character. This is a fair approximation to the

No. 626] GIGANTISM IN NICOTIANA TABACUM 225

expected ratio 1820/4— 455, if the Mammoth habit be-
haved as a simple Mendelian character in contrast with
the normal blossoming habit.

From these data it would appear that the Mammoth
character is recessive in its inheritance and reappears in
the F, generation in numbers approximating closely the
expected ratio for a simple Mendelian recessive.

THE Or AND BEHAVIOR OF A New Mammoru TYPE OF
TOBACCO IN A LINE DESCENDING FROM A SPEctEs HYBRID

In an earlier paragraph it has been mentioned that the
Maryland Narrowleaf Mammoth and a Burley Mammoth
appeared in the F, generation of certain crosses. In the
writer’s experience a giant type appeared in third gen-
eration plants descending from a species cross.

In 1914 the blossoms of a first generation plant of the
cross Connecticut Broadleaf (pink) 2 X Giant Red flow-
ering (carmine) ¢ were pollinated with the pollen of
Nicotiana sylvestris (white). Although first generation
plants of crosses between the species N. tabacum and N.
sylvestris are likely to be sterile, or nearly so, consider-
able fertile seed were obtained from F, generation of this
particular cross. In the second generation there was a
noticeable segregation into plants with pink, white and
carmine blossoms. The size and shape of the blossoms
of the plants of the F, generation were also very variable
and various abnormalities were noted. Some plants were
completely self-sterile and others produced blossoms with
supernumerary petals. A number of plants producing the
largest and finest carmine-colored blossoms were selected
for further inheritance studies. The progenies of two of
these mother plants, nos. 9 and 12, were grown in the field
at Arlington, Va., during the season of 1916.

The mother plant, no. 9, proved to be heterozygous,
breaking up into carmines and pinks, approximating the
theoretical ratio of three carmines to one pink. All the
plants of this line were normal in size and habit of growth.

3 The so-called Giant Red flowering tobacco sold by seedsmen for orna-
mental purposes, is only a variety of N. tabacum with deep carmine blossoms.

226 THE AMERICAN NATURALIST [Vou. LIII

The sister plant, no. 12, which proved to be homozygous
for carmine, behaved differently, giving rise to a progeny
of plants which were very variable in height.* A number
of these plants appeared to possess the Mammoth habit
of indeterminate growth and gave no evidence of blossom-
ing. On October 26, 1916, the heights of the plants, all
of which had blossomed except those of Mammoth habit
of growth, were as follows:

TABLE III.
HEIGHTS OF THE PLANTS IN THE PROGENY oF SISTER PLANT No. 12

Height classes
3 to 5 ft. 5 to7 7 to 9 ft.
Number in class.... 12 (blossomed) 16 hloas) 3 (Mammoth)

The shortest plants in this progeny were first to blos-
som and produced an average of only 20 to 25 leaves, in-
cluding the first bald sucker. Other plants of intermedi-
ate heights blossomed considerably later and produced an
average of 35 to 40 leaves, including the first bald sucker.
Those plants of Mammoth habit of growth which showed
no indications of blossoming had produced considerably
more than 40 leaves.

Two of these Mammoth plants, nos. 12 (a) and 12 (b),
each seven feet in height, were transplanted in the green-
house October 21 without cutting them back. Both plants
blossomed December 8, producing carmine blossoms.
Plant no. 12 (a) had produced 70 to 75 leaves, not includ-
ing many bract-like leaves below the flowerhead. Plant
no. 12 (b) produced 60 to 65 leaves, including all small
ones below the flowerhead.

In addition to these two Mammoth plants the seed of
several of the taller sister plants, nos. 12 (c) and 12 (d),
in class 2, which had blossomed late, producing 35 to 40
leaves, were saved separately. The progenies of all were
grown in the field at Arlington Farm, Va., in 1917. A

+The leaves of the mother plant no. 12 were characterized by coarse,
thick, broad and rounded blades abruptly contracted at the base to a long,
almost naked or slightly winged petiole. This striking type of leaf has re-
mained constant in the progeny of no. 12, and also in the progenies of no.
12 (a), 12 (b), 12 (e) and 12 (d), descending from this mother plant.

No. 626] GIGANTISM IN NICOTIANA TABACUM 227

total of 60 plants was grown from the Mammoth mother
plant, no. 12 (a), all of which were of Mammoth type, with
an average height of seven to seven and a half feet. On
September 11 a few of the tallest plants were eight feet
in height. On this date an average of 50 to 55 leaves had
been produced and none showed any evidence of blossom-
ing. A progeny of 60 plants (see row 38A, 1917) was
also grown from the Mammoth mother plant, no. 12 (b).
On September 11 these plants averaged six and a half to
seven feet in height and resembled the progeny of no. 12
(a) in all respects except that they were not quite as tall.

From the mother plant, no. 12 (e), which was one of
the late blossoming plants, producing an average of 35
to 40 leaves, 49 plants were grown. On September 13
the heights of 48/of these plants and their blossoming
habits were noted as follows:

TABLE IV

HEIGHTS OF 48 PLANTS IN THE PROGENY oF MoTHER PLANT No. 12 (c) SE-
LECTED FROM CLASS 2, oF TABLE IT

Height of classes
7 to 9 ft. 1 to 11 ft
act Mamm. a Mamm. Pye Mamm,
Number in class .. 2 12

0

The height of one ee which ae was not obtained ae is not in-
cluded in the table.

In this progeny of 49 plants it is evident that 14 plants
possessed Mammoth characteristics of continuous growth
and showed no evidence of blossoming, while 35 plants,
some of which were of giant stature, blossomed. From
the late blossoming mother plant, no. 12 (d), a progeny
of 48 plants was grown. The heights of 42 of these plants
were also measured on September 13 and their blossoming
habits noted as follows:

TABLE V

FREQUENCY DISTRIBUTION OF HEIGHTS OF 42 PLANTS IN PROGENY OF MOTHER
PLANT No. 12 (d) SELECTED FROM Cuass 2 or TABLE III

Height of class
3 to 5 ft. 5 to 7 ft. 7 to 9 ft 9 to 11 ft.
Normal Mamm, Normal Mamm. Normal Mamm Normal oN
Number in elass .... 0 1 17 0 19 0

Six other plants were grown in this progeny which are not included in w
table since their heights were not obtained. All blossomed, however.

228 THE AMERICAN NATURALIST [Von. LIL

In addition to these individual progenies of the sister
plants, nos. 12 (a), (b), (c) and (d), selected from the
progeny of the mother plant, no. 12, in 1916, a mixed lot
of seed was harvested from several other sister plants
which had blossomed. Fifty-six plants were grown from
this mixed lot of seed, all averaging six to six and a half
feet in height, and all blossoming. In this lot of plants
there were no indications of Mammoth types and so far
as could be determined with the eye, no intermediate
forms were present.

From the inheritance behavior of the sister plants, nos.
12 (a), (b), (c) and (d), it is evident that pure Mammoth
types, breeding true, and intermediate inconstant types
appeared simultaneously in the progeny of the original
mother plant, no. 12. These intermediate plants behaved
as hybrid forms, in that they gave rise in their progeny
to a certain percentage of typical Mammoth, non-blossom-
ing types. Since the progenies of the two sister plants,
nos. 12 (c) and (d), were handled under similar condi-
tions from the time the seed were sown, it is evident that
the mother plant, no. 12 (c), yielding 14 Mammoth plants
in a total of 49 plants, was considerably more prolific in
Mammoth individuals than the sister plant, no. 12 (d),
which yielded only two Mammoth individuals in a total of
` 48 plants.

It is of interest to note that Lodewijks (1911) in work-
ing with tobacco in Java, has observed the occurrence of
Mammoth types which breed true and also intermediate
or inconstant races which break up into Mammoth or
Giant forms approximating the theoretical Mendelian
ratio of 25 per cent.

Lodewijks regards these inconstant races as hybrid
mutations and states the results of his investigations as
follows, a translation of which will also be given:

TRANSLATION
I. Occasionally giant plants which breed true to type occur in Vor-
stenland tobacco.
II. Evidently giant intermediate races also oceur.

No. 626] GIGANTISM IN NICOTIANA TABACUM 229

III. In my experiments I obtained either an atavist of an inconstant
intermediate race or a hybrid-giant.

IV. As none of the giant plants in my experiments have reached
the flowering stage, it is not certain which of the two mentioned possi-
bilities is the chief. It would seem to be the latter, however, as seed
of the few-leaved mother plant of the second generation produced ex-
elusively plants while seed of the many-leaved plant produced nearly
25% giant and many-leaved and few-leaved plants.

V. It is probable, therefore, that a second instance is present of a
mutation arising as a hybrid.

Honing (1914), in his studies of the aberrant types
occurring in Sumatra and Java tobacco, states that in
some instances 100 per cent. of the progeny of normal
plants were of the Mammoth type. According to Honing
even the Mammoth plants were not always constant in
their inheritance, and intermediate races were also
present.

From Lodewijk’s observations in J ava, and the writer’s
observations at Arlington Farm, Va., it is evident that
intermediate races, as well as Mammoth types which
breed true, may appear in a progeny. Concerning the
actual mode of origin of these intermediate and Mam-
` moth races nothing definite is known. Hayes and Bein-
hart (1914), speaking of the origin of a Mammoth Cuban
type in Connecticut in 1912, say:

This mutation must have taken place after fertilization, i. e., after
the union of the male and female reproductive cells. If the aakh
had taken place in either the male or female cell before fertilization,
the mutant would have been a first generation hybrid, and would have
given a variable progeny the following season.

They assume that if one gamete alone were affected, a
progeny of hybrid character would have resulted, but if
we assume that one gamete can become so affected, it is
quite as reasonable to assume that both may sometime
be changed in the same manner. If such were the case,
Mammoth plants breeding true to this indeterminate
habit of growth would be expected.

If, as Lodewijk finds, intermediate races behave as true
Mendelian hybrids, producing the theoretical ratio of 25

230 THE AMERICAN NATURALIST (Vou. LIH

per cent. true Mammoth plants which breed true, there
is strong reason to believe that the change responsible
for Mammoth habit of growth has affected one gamete
only. If this gamete unites with a normal gamete, then
the simple Mendelian ratio would follow, just as in the
case of an artificial cross between gametes produced by
a Mammoth plant and those of a normal plant. In the
one case a portion or all the gametes bearing the Mam-
moth character are produced by a normal plant. In the
other case, Mammoth plants themselves produce gametes
with potential Mammoth characters. In the experience
` of Honing, normal plants have even produced progenies
containing 100 per cent. mammoth plants. This behavior
would indicate that all the gametes produced by a mother
plant may sometimes become modified to express the
Mammoth habit of growth. Although Honing has ob-
served the complete disappearance of a line of Mammoth
plants which gave rise to progenies of blossoming plants,
this behavior has not been definitely observed in this
country except as a response to obscure environmental
conditions. It is possible that the behavior of Honing’s
inconstant Mammoths is of this nature rather than an
internal gametic change, permanently affecting the hered-
ity of the Mammoth feature. Until this question is more
definitely settled, Honing’s inconstant Mammoth can not
be disposed of.

Since inconstant, intermediate plants behaving as Men-
delian hybrids with respect to Mammoth character, and
sister plants of pure type are known to arise suddenly in
the same progeny, there is reason to believe that the
change responsible for Mammoth behavior may affect one
or both gametes, as the case may be. This inconstant be- ©
havior of these mutant hybrids is particularly significant
since it appears in every way similar to the actual be-
havior of a controlled cross between a Mammoth and a
normal plant. Of course, if it is possible for one or more
gametes produced by a normal plant to become so modi-
fied as to originate a hybrid-mutant or a pure line mutant,

No. 626] GIGANTISM IN NICOTIANA TABACUM 231

then it is quite as probable that all the gametes in a single
blossom, or the gametes produced by all the blossoms of
a normal plant, may become so modified. Honing’s ob-
servations at least would indicate that this does occur.
In those instances where an occasional Mammoth ap-
pears in the progeny of a normal plant, it is usually as-
sumed that the change responsible for the Mammoth char-
acter was associated in some way directly with the ga-
metes themselves. In those instances where many or even
all the plants in the progeny of a normal plant produce
Mammoths, the question becomes more involved and dif-
ficult of interpretation. It is very difficult to see how all
the gametes of a normal plant can become simultaneously
modified to produce by their union Mammoth plants, un-
less we assume that the change takes place at some stage
preceding the development of the gametes. Should the
change take place in a mother cell of the anther preceding
tetrad formation, i. e., by the addition or subtraction of
some factor in the chromosome material, it is reasonable
to suppose that the four pollen grains resulting from the
division of this mother cell may be similarly affected, and
bear. the Mammoth character. It is possible, however,
that the change may take place very much earlier, so that
a part or even all the sporogenous cells will be affected.
If this condition occurred, it is easy to see how great num-
bers or even all the pollen grains arising from their di-
vision would bear the Mammoth character. Since the
development of the megasporangium is in every way par-
allel to the development of the microsporangium or an-
ther, similar changes would affect one or more egg-cells,
depending upon whether the change responsible for Mam-
moth character took place immediately in the egg-cell it-
self, in the mother cells, or very much earlier, so that all
the sporogenous cells, and hence all the egg-cells arising
from them, are affected. Such changes affecting great
numbers or all the gametes in a single flower, or even
in the entire flower head itself, would produce the phe-
nomenon of a more or less complete acquirement of Mam-

232 THE AMERICAN NATURALIST (Von. LIIL

moth character in the progeny of a normal plant. It may
be stated here that East (1917) has offered the same sug-
gestion concerning the origin of variations in cell-divi-
sions preceding the formation of the gametes themselves.

Tue Propuction or New MammotH Forms By
HYBRIDIZATION

Two Mammoth types of tobacco are now grown com-
mercially in the United States, the Maryland Narrowleaf
Mammoth in Maryland, and to a lesser extent the Stew-
art Cuban in the Connecticut Valley. Promising Mam-
moth types have also originated in Havana Seed tobacco
in Connecticut. Beinhart (1918, however, in a brief dis-
cussion of the occurrence of Mammoth types in the Con-
necticut Valley, states that practical methods of seed pro-
duction and special cultural methods must be worked out
before the Stewart Cuban Mammoth and the Havana Seed
Mammoth can be successfully grown on a commercial
seale. Although these Mammoths originated spontane-
ously from commercial types, there is every reason to
believe that valuable new types can be secured by cross-
ing with the ordinary commercial types of tobacco.
Since in crosses with ordinary varieties gigantism is re-
cessive in its inheritance, the problem of producing new |
giant types by hybridization and recombination has not
been difficult. Several Mammoth types have already
been secured in crosses with Connecticut Broadleaf, Little
Dutch and White Burley. If by this means it is possible
to combine the habit of gigantism, which insures greatly
increased yields, with the desirable quality characteris-
tics of ordinary varieties, very valuable commercial types
can be obtained.

SuMMARY

1. Gigantism has occurred in several different commer-
cial varieties of tobacco, including Maryland types, Cu-
ban, Connecticut Havana and Sumatra. It has also been
associated with certain varietal crosses and species
Crosses.

No. 626] GIGANTISM IN NICOTIANA TABACUM 233

2. Not only giant or mammoth types which breed true,
but intermediate or hybrid types occur spontaneously
which subsequently give rise to a greater or less propor-
tion of mammoth forms.

3. In crosses with normal varieties the mammoth char-
acter is recessive, and F, plants invariably blossom. The
F, plants average a somewhat higher leaf number than
the normal parent which entered into the cross.

4. In the F, generation mammoth plants occur in pro-
portions approaching the theoretical ratio of 25 per cent.
obtaining in a single Mendelian cross involving two con-
trasted unit characters.

LITERATURE CITED
Beinhart, E. G.
1918. Uncle Sam and His Colleagues in the Connecticut Valley. To-
bacco, New York, 66: Sept. 26, pp. 35-39.
East, E. M.
1917. The Bearing of Some General Biological Facts on ee
tion. THE AMERICAN NATURALIST, 51: March, pp. 129-
Garner, W. W.
1912. Some Observations on Tobacco Breeding. Rept. Amer. Breeders’
Assn., 8: pp. 458-468.

H. K.
1915. Tobaceo Mutations, Jour. Heredity, 6: No. 2, Feb., pp. 73-78.
Hayes, H. K., and Beinhart, E. G.
1914, Mutation in Tobacco. Science, N. S., 39: No. 922, pp. 34-35.
Honing, J. A.
1914. Deli-Tabak, een Mengel van Rassen die in Bladbreedte en Aantal
Bladeren Verschillen.’’ nc ate van het Proefstation te
Medan, 8: No. 6, pp. 155-
Hunger, P.W. T.
1905. Untersuchungen und Betrachtungen über die Mosaikkrankheit
der Tabakspflanze. Te Fd Pflanzenkrankheiten, 5:
No. 5, pp. 257-311. (See page 277.)
Lodewijks, J. A.
1911. Erblichkeitsversuche mit Tabak. ss oe T Induktive
Abstammungs-V ererbungslehre, 5: pp.

THE MENDELIAN BEHAVIOR OF AUREA CHAR-
ACTER IN A CROSS BETWEEN TWO
VARIETIES OF NICOTIANA
RUSTICA

H. A. ALLARD
OFFICE OF TOBACCO AND PLANT NUTRITION INVESTIGATIONS, BUREAU
oF Puant Inpustry, U. S. DEPT. OF AGRICULTURE,
WASHINGTON, D. C.

INTRODUCTION

THE species of tobacco Nicotiana rustica comprises a
number of more or less distinct varieties. One of the
more characteristic varieties which was received from
Russia (S. P. I. 35080) is a light, yellowish-green type
with distinctly white stems and midribs. In these re-
spects this type of N. rustica resembles the well-known
White Burley variety of N. tabacum. According to
Splendore,' who has described it in. detail, this white-
stemmed variety of N. rustica is grown commercially in
- Russia (Makorka, Bakoun, Kolmak, Tseco, ete.) as a pipe
and cigarette tobacco. In this variety of N. rustica, the
stems of young plants—especially if they have been
somewhat etiolated by crowding—are almost snow white.
A cross section of the stems of such plants one month
old reveals the fact that this whiteness is not merely su-
perficial, but extends entirely through the stems, whereas
in green varieties of N. rustica the internal structure of
the stems is green throughout. The cotyledons are de-
cidedly chlorotic and the leaves have a pale yellowish-
green, chlorotic appearance which becomes more marked
as the plants approach maturity. As a matter of con-
venience, the writer has applied the term ‘‘aurea’’ to
this peculiar, varietal form of chlorosis.?

1 Splendore, A., ‘‘Due Particolore Forme Di N. Rustica Brasilia Chwit-
zent e Kapa Magiara,’’ Boll. Tech. Della Colt. Dei Tabachi del R. Inst.
sperimentale, Scafati (Salerno), XI, No. 2, 1912.

2 This type of aurea appears to be quite distinct from the type of aurea
deseribed by Lodewijks as having occurred suddenly in plants of N. taba-

No. 626] MENDELIAN BEHAVIOR 235

DOMINANCE OF GREEN Puant Coror in F! PLANTS OF
ROSSES

In 1914 the writer made reciprocal crosses of this
white-stemmed, chlorotic aurea type of N. rustica, with
green-stemmed, green-leaved type, and several hundred
F! plants were grown in the field at Arlington, Va., dur-
ing the season of 1915. All the F' plants were green in
color, whichever type was used. as the seed-bearing par-
ent. It was at once evident that the white-stemmed,
chlorotic, aurea character behaved as a simple recessive
to normal greenness of stem and leaf. To determine
more fully the Mendelian behavior of this cross an anal-
ysis of the F? and F? generations was made.

SEGREGATION IN F2 PLANTS

In the F? generation, green- and white-stemmed aurea
plants appeared. So distinct is the white-stemmed re-
cessive that four or five weeks after germination, the
young plants can be readily distinguished from the
green-stemmed types. This made the growing and han-
dling of large numbers of plants a comparatively easy
matter, since it was only necessary to grow them to the
size of small seedlings and obtain counts when they were
four or five weeks old. In the following table an analysis
of 25,000 F+ plants is shown

From this data it is evident that the recessive white-
stemmed aurea type of rustica appeared in numbers ap-
proximating very closely the theoretical Mendelian ratio
of 25 per cent., since in a population of 25,000 plants,
24.31 per cent. were of the white-stemmed, aurea type.
cum in Java. He found that this type of aurea was inconstant in its in-
heritance, since awrea mother plants always gave progenies consisting of

een and aurea plants. In crosses between aurea and green plants the F1
asa always included aurea and green plants. From the inconstant

eritance of this character he concludes that this aurea form originated

ant with an essentially hybrid constitution. See Lodewijks, J. A.,

‘*Erblichkeitsversuche mit Tabak,’’ Zeitschr. fiir Induktive Abstammungs.
Vererbungslehre, Vol. 5, 1911, pp. 139-172.

236

THE AMERICAN NATURALIST

TABLE I
RATIOS OF GREEN-STEMMED AND RECESSIVE WHITE-STEMMED AUREA PLANTS
IN THE F1 GENERATION OF THE CROSS 35080 (WHITE-STEMMED)

[Von. LIII

X No. 1 FROM INDIA (GREEN-STEMMED) ĝ

Date of Count Total Number | Number of _ | white-stemmed | White-stemmed

Counted | Green-stemmed aurea aurea

June 19, 1918... es. 4,188 3,178 1,010 24.1
May YO; 3918 os a 1,167 ' 887 280 23.9
May 17; 3916) 352 cco 417 308 109 24.9
June 24, 1918... .00..05. 1,955 1,476 479 24.5
June 26, LOIBA n 279 231 48 17.2
July TOIR eo, 1,072 818 254 23.6
SUS LO, LOLO.. Pees se 747 594 153 20.4
Jely iPr Aoi eeii 1,246 930 316 25.8
Sept. 3, 1918 u: a 2,114 1,597 517 24.4
Oct. 18, TUF oct. ies 2,253 1,707 546 24.2
Oet. 21; 101s... 2,673 2,035 638 23.8
Oct 24 TOR ore 4,556 3,400 1,156 25.3
Qot: 20; IVIR, she es 2,333 1,760 573 24.5

TOLA oy a iS 25,000 18,921 6.079 24.31

BEHAVIOR oF F? GREEN PLANTS AND WHITE-STEMMED

EXTRACTED aurea RECESSIVE

Of twenty-eight F? green plants selected at random the
character of the inheritance in the progenies of those
showing segregation was noted as follows:

TABLE II
RATIOS OF GREEN-STEMMED AND WHITE-STEMMED AUREA PLANTS APPEARING

IN THE PROGENIES OF HETEROZYGOUS GREEN-STEMMED PLANTS IN
THE F2 GENERATION
Number of Mother Total Number of gps esate of Green- Number of White-
Plant Progeny Counted itemmed Plants stemmed aurea Plants

1 414 300 114

4 401 ~ 300 101

5 462 345 Liz

6 392 300 92

7 887 649 238

10 403 304 99

11 395 300 95

12 386 300 86

14 399 300 99

17. 390 300 90

21 410 310 100

23 397 300 97

24 660 503 157

27 995 747 248

28 739 561 178

33 393 300 93
ORRIN os Sees 8,123 6,119 2,004

No. 626] MENDELIAN BEHAVIOR 237

The green F? individuals Nos. 2, 3, 9, 13, 15, 16, 18, 19,
20, 25, 29, 30 were homozygous for greenness and gave
pure green progenies. A progeny of several thousand
plants was grown from each individual.

Of 8,123 plants descending from heterozygous green
individuals analyzed in Table II, 2,004 or 24.6 per cent.
were white-stemmed recessives. It is evident that these
figures for the extracted recessives also approach very
closely the theoretical 25 per cent. Mendelian ratio which
obtains for contrasted characters in simple hybrids.
This ratio of 24.6 per cent. extracted recessives of the
aurea type descending from green heterozygous F? indi-
viduals, is very close to the ratio 24.3 per cent. obtained
in a count of 25,000 individuals descending from F!
plants. Since 12 of the 28 green F? plants tested were
homozygous for greenness and gave all green progenies,
it is evident that these were extracted dominants.

The progenies of 20 extracted white-stemmed aurea
recessives of the F? generation were also studied. Sev-
eral thousand plants were grown from each of the 20
individuals, and all Be ho homozygous for the aurea
character, ain.

BEHAVIOR OF Back CROSSES

First generation plants of the original cross No. 35080
(white-stemmed aurea) 2? X No. 1 from India (green-
stemmed Rustica) 3 were now crossed with the parent
green-stemmed and white-stemmed aurea types.

In the back cross with the recessive white-stemmed
aurea parent 591 plants were obtained of which 303 were
green-stemmed individuals, and 288 were aurea. These
figures approach the theoretical 1:1 ratio which may be
expected in such crosses.

In the back cross with the dominant ereen-stemmed
type, 280 plants were obtained, all of which were green-
stemmed.

From these results obtained with the cross between the
green-stemmed types of N. rustica and the distinctive

238 THE AMERICAN NATURALIST [Von. LIII

white-stemmed aurea type, it is evident that we are deal-
ing with a clear-cut instance of Mendelian behavior, in
which greenness of stem and leaf is contrasted with the
character of white stems and a yellowish, chlorotic ap-
pearance of the leaves. Since these characteristics are
readily distinguished in plants in the seedling stage, only
five or six weeks after germination, this cross is espe-
cially favorable for the demonstration of simple Men-
delian behavior in all its phases. The technique of
crossing is simple, and many thousands of seedlings may
be grown in a comparatively small area in a short time.

SuMMARY

In crosses between a distinctive white-stemmed aurea
type and green-stemmed type of N. rustica the following
Mendelian relations were found:

In Ft plants the white-stemmed aurea type is recessive
to the green-stemmed type.

F? plants segregate into green-stemmed and white-
stemmed aurea plants. Approximately 25 per cent. of
the plants are aurea recessive. Some of the green plants
are homozygous for greenness of stems, etc., and some
are heterozygous, again segregating into green and
white-stemmed aurea types with the same ratios ob-
tainéd in the F! generation. The extracted aurea re-
cessives of the F? generation are homozygous with re-
spect to the character of white stems, ete., peculiar to
this type.

In back crosses between a heterozygous F* plant and
the dominant green-stemmed type, the progeny consists
of 100 per cent. green-stemmed plants.

In back crosses with the recessive aurea type, the
progeny consists of green-stemmed and white-stemmed
aurea plants in ei ating ane the expected ratio of
1 to 1.

SOME FACTOR RELATIONS IN MAIZE WITH
REFERENCE TO LINKAGE

D. F. JONES AND C. A. GALLASTEGUI

CONNECTICUT AGRICULTURAL EXPERIMENT STATION

In view of the many distinct Mendelizing characters
known in maize (Zea mays L.) it has been rather sur-
prising that so few eases of linkage have been reported in
this plant up to the present time. The number of chro-
mosome pairs, about ten, is not large for plants and about
twenty distinct contrasting factors are known of which
the inheritance can be easily followed and about as many
more which offer some difficulty in following in transmis-
sion but which can be used more or less satisfactorily in
carrying on experiments on linkage. The writers have
made no systematic search for cases of linkage in maize,
but having found, almost accidentally, what seems to be
a fairly good ease of linkage between the tunicate factor
which determines the production or inhibition of the
glumes covering the seeds and the factor for starchy or
sweet endosperm, the results are reported here in the
hope that they may be of use to others who may be pursu-
ing investigations along this line.

Collins and Kempton (1911) were the first to record a
case of linkage in maize. Their results involved the re-
lation of endosperm texture, as contrasted in our ordi-
nary starchy varieties with the waxy condition found in
Chinese varieties, to the color of the aleurone layer. They
did not determine which of the several factors concerned
with aleurone color was involved in this linkage. More
= recently Bregger (1918) has given additional proof of
this case of linkage. He has determined the amount of
crossing-over and has also shown that it is the C aleurone
factor (East and Hayes, 1911; Emerson, 1918) which is
the one involved. At about the same time Lindstrom
(1917) reported the second case of linkage, that of one of

T ' 289

240 THE AMERICAN NATURALIST [Von LIT |

the factors of chlorophyll color G, with another aleurone
color factor, this time the R factor, which in the presence
of a suitable basic factorial combination produces red
color in the aleurone cells. More recently Lindstrom
(1918) has found another chlorophyll factor L linked with
R and G. L is completely linked with R and both show
about the same amount of breaks in the linkage with G.
This makes the first group of three factors so far reported
in maize.

LINKAGE BETWEEN TUNICATE EAR AND STARCHY-SWEET
ENDOSPERM FACTORS

The curious type of maize, known generally in this
country as pod corn (Zea mays tunicata Sturtevant) is
considered by Collins (1917) not to be a pure type, but a
heterozygous condition somewhat analogous to fhe blue
Andalusian fowl. When selfed seed of the typical podded
ears are planted Collins finds that three types of plants
are produced: one type like the typical podded parent;
one with normal ears without the enclosing glumes; and
one anomalous type of a plant which does not produce
seed in lateral inflorescences, but in perfect flowers in the
tassels. On these last plants lateral inflorescences with
much elongated glumes are produced, but are sterile. Al!
these three types have been secured in about the ratio of
1:2:1 as expected on the assumption that a single Men-
delizing difference is involved and the heterozygote is
distinguishable from both homozygotes.

Our own rather limited experience with this type of
maize confirms Collins’s conclusions. In 1915 seed of a
typical podded ear was planted (there was no record
whether it had been selfed or not). All three types which
Collins described later were obtained. A number of typi-
cal podded plants were self-pollinated and grown the
next year in the hope of getting a pure podded strain. At
that time no thought was given to the possibility of its
being a heterozygous type. The plants with seeds in the
tassels were thought to be extreme variations from the

No. 626] FACTOR RELATIONS IN MAIZE 241

usual type. Nine selfed ears were obtained and grown
the following year and all gave some plants with podless
ears and others with seeds in the tassels as well as plants
of the typical pod type. No record was made of the num-
bers in each class, but it was noted as rather surprising
that all of the nine ears gave some normal non-pod plants.

An attempt was made to self-pollinate some of the
plants with the peculiar terminal inflorescences which
were easily recognized as the type which produced seed,
but no seed was obtained where they were enclosed in a
bag. Very little good pollen is produced in these tassels
and probably all or most of the seeds which are produced
on open-pollinated plants result from crossing with for-
eign pollen. One tassel with a number of such open-pol-
linated seeds was saved and the seed planted. No normal
non-pod ears were obtained. Most of the ears were of
- the typical pod type or half-tunicate as named by Collins.
All of these results bear out the assumption of Collins
that the podded maize considered by Sturtevant (1899)
as a separate species and stated by him to have been
known for 300 years is not a constant type and has little
more claim to specific rank than the blue Andalusian
fowl.

One of the half-tunicate ears produced from the open-
pollinated seed of the perfect flowered segregate was self-
pollinated, and when examined this ear was found to have
segregated into starchy and sweet seeds, showing that the
plant which had furnished the pollen had sweet seeds.
Since all the podded maize which had been grown up to

_that time was starchy and all of the sweet maize was non-
podded, the cross involved the tunicate character and
starchy endosperm from the female parent, and non-
tunicate, sweet endosperm from the male. The starchy
and sweet seeds were planted separately. There were 173
of the starchy and forty-three of the sweet seeds. Nota
perfect 3:1 ratio, but reasonably close. All of the seeds
were planted, but since not all of each type produced a
mature plant, it is legitimate to correct the observed re-

242 THE AMERICAN NATURALIST [Vou. LIII

sults according to the theoretical starchy-sweet ratio.
The results obtained and corrected in this way are as
follows:

Starchy Starchy | Sweet | Sweet
Tunicate Non-tunicate | Tunicate | Non-tunicate
Pend aa bt ais 4 a Eor
Uorretted: ae ae kei | 108.0 3.8 | 8.2 | 29.1
Starchy-sweet ratio | |
Ptpacta ceren ie; | 105.8 6.0 | 6.0 | 31.3
11:1:1:11 gametic ratio | | |

The numbers are small, but the distribution obtained is
clearly different from a 9:3:3:1 ratio. The agreement
with the nearest theoretical results, assuming linkage, is
close (P=.615). The per cent. of crossing-over, 8.3, in-
dicated by these figures is low. In the other cases of
linkage reported, the percentages of crossing-over were
much higher, 25.7 per cent. in the waxy endosperm-
aleurone color combinations and 20 per cent. in the
aleurone color-chlorophyll color combination with the
exception of the one case where complete linkage has so
far been found.

In making the classification all of the plants which
showed the tunicate character, whether of the half-tuni-
cate or full-tunicate type, were classed as tunicate, as con-
trasted to the normal plants. Segregation was clear be-
tween these two classes and there was little possibility of
confusion even when the ears were immature. On the
other hand, it was not always easy to distinguish full-
tunicate from half-tunicate plants, as the tassels of the
former class do not always produce seed, and the ears,
which are quite characteristic when fully developed, are
not so distinct when immature, and many of these plants
were late in maturing. Any error of classification here
does not affect the linkage results, however. There is one
source of error in that the plants suckered profusely;
many of these bore ears and tassels and were difficult to
distinguish from the main stalk. The plants were grown
in hills, and when classifying them it was not always pos-

No. 626] FACTOR RELATIONS IN MAIZE 243

sible to tell which was plant and which was sucker, so
that the same plants may have been included in the count
more than once.

The figures for the segregation with respect to the tuni-
cate character together with the figures from a similar
ear, which instead of segregating starchy and sweet seg-
regated for yellow and white endosperm, are aier as
follows:

| Normal Half-tunicate | Full-tunicate
| 25 |
EE OF |
Ear 1 { her eee ey 4 a ei
nadra garrena E y P es 10 19 14
Ear 2 { Citing a A es. E 36 58 29
a OS | 75 | 153 87
79

TEI sn fate Hh. inn pines donee e8 | 79 | 158
1:2:1 ratio |

With regard to this second ear, which was similar to
the first except that it was crossed with yellow, starchy,
non-tunicate instead of white, sweet maize, it is to be
noted that there is no indication of linkage between the
factors for tunicate ear and yellow endosperm. The fig-
ures obtained compared to the expectancy are given
herewith

Yellow Yellow | White White
Tunicate Non-tunicate | Tunicate Non-tunicate
A E od | cae: 87 36 oe 10
ge SS 88.1 36.4 = | 31.9 9.7
eniveahiad ratio |
Me ee 93.4 31.1 | 31.1 10.4
9:3:3:1 ratio | |
P= .739 | |

EVIDENCE FOR LINKAGE BETWEEN ALEURONE COLOR
FACTORS

Another case of linkage is suggested by the results of
East and Hayes (1911) in the inheritance of aleurone
color. From crosses of colorless aleurone by purple they
obtained marked deviations from the expected ratios
which they could not account for. At that time the first

244 THE AMERICAN NATURALIST [Vok LIIT

cases of gametic coupling had just been published by
Bateson and Punnett and the subject of linkage was not
well understood nor its true significance realized. These
writers considered the possibility of gametic coupling as
a disturbing factor, but came to the conclusion that this
phenomenon could not be concerned in their aberrant
results.

One of the crosses studied involved two aleurone fac-
tors, the basic color factor C and the factor P (Emerson’s
Pr factor), with R present coming from both parents.
This cross was expected to give aratio of 9 purples: 3 red:
4 non-colored, but actually showed a large excess of pur-
ples and deficiency of reds. East and Hayes considered
the possibility of linkage between the P and R factors,
but since R, according to their theory, was homozygous,
crossing-over between these two factors would make no
visible difference in the F, results. On the other hand,
since PC entered the cross from one side and pe from the
other, the cross-over class pC, if there is linkage, would
be red because of the presence of R. Hence any possibil-
ity of linkage should be looked for between the P and C
factors. Such a situation would account for an excess of
purples and a deficiency of reds in the cross under con-
sideration. Another cross involving, in addition to the P
and C factors, a color inhibiting factor likewise showed
an excess of purples and a deficiency of reds.

Since it is always rather difficult to prove linkage from
F, distributions alone, in this case it would be even more
difficult because only one of the cross-over classes, if such
it is, can be distinguished. The data of East and Hayes,
as far as numbers go, do not agree with expectation from
linkage with any amount of crossing-over, and since other
crosses involving the same factors have been reported
which seem to show independence, it is doubtful whether
or not linkage really exists in respect to these two factors.
Tt is more probable that the deviations from theory are to
be looked for in either incomplete analysis of the factor
relations or faulty classification of the seeds. Red seeds

*

No. 626] FACTOR RELATIONS IN MAIZE 245

gradate somewhat into purple and there may be a ten-
dency to include reds among purples. If this were the
case, however, wrongly classified purple seeds should
sometimes give all red progeny or red and white progeny
in the next generation. East and Hayes found no cases
of this kind. The possibility of linkage between the P
and C factors should be kept in mind until this point can
be definitely settled.

OTHER FACTORIAL RELATIONS

Looking over East and Hayes’s data for other cases of
linkage or independence of factors there seems to be good:
evidence that the C aleurone and R aleurone factors are
not linked, and also that the factor for sweet endosperm
is not linked with either the R or P aleurone factors. In
a factorial analysis of the characters of an organism with
reference to linkage it is just as important to know the
cases where no linkage is shown as those cases where it is
shown. Collins and Kempton (1913) give data which in-
dicate independence between sweet and waxy endosperm
factors and in another paper (1917) independence be-
tween the tunicata and ramosa factors. East (1910) gives
data which indicate that the two factors for yellow endo-
sperm color are not linked with each other and it is quite
probable that both of them are independent of the factor
for sweet endosperm.

With this evidence we can attempt a beginning at an
analysis of the factorial relations in maize. Three inde-
pendent groups of factors can be tentatively proposed as
follows:

Group I | Group II | Group III
Ww Endosperm Gg Chlorophyll | Ss Endosperm
C urone Li Chlorophyll | Tt Tunicate
_ Pp Aleurone (?) | !

The fact of no linkage between the Cc aleurone and the
Rr aleurone color factors separates groups I and II. No
linkage between Ww and Ss endosperm factors separates

246 THE AMERICAN NATURALIST [Vou. LIII

groups I and III. No linkage between the Ss endosperm
and Rr aleurone factors separate groups II and ITI.

Since the number of known factor differences in maize
is already some three or four times the number of chro-
mosomes, more definite knowledge of the behavior of all
these factors in relation to each other will be awaited with
interest. Especially since maize is one of the best mate-
rials from the plant side to which the chromosome hy-
pothesis, as worked out in Drosophila, can look for con-
tradiction or support.

LITERATURE CITED
Bregger, T.
1918. Linkage in Maize: The © Aleurone Factor and Waxy Endo-
sperm. AMER. Nat., 52: 57-61.
Collins, G. N.
1917. Hybris of Zea Ramosa and Zea Tunicata. Jour. Agric. Re-
rch, 9: 383-395, pls. 13-21.
Collins, G. x. d Kempton, J. H.
1911. Tnheritanc of Waxy Endosperm in ng Ve of Chinese Maize.
IV° Conf. Internat. de Tapo ge 357.
1913. ee of Waxy Endosper aa with Sweet Corn.
r. Plant. Ind. Cir. 120, pp. oh i.
East, E. M.
1910. A Mendelian Interpretation of Variation that is Apparently
Continuous. AM. NAT., 44:
East, E. M., and Hayes, H. K.
1911. Teei in Maize. Conn, Agric. Exp. Sta. Bull., 167, pp. 142.
Emerson, R, A.
1918. i Ta Pair of Factors, Aa, for Aleurone Color in Maize, and
s Relation to the Ce and Rr Pairs. Cornell Univ. Agric. Exp.
od Memoir 16, pp. 225-289.
Lindstrom, E. W.
1917, ror) in Maize: Aleurone and Chlorophyll Factors. AMER.

1918. Gilera Tuheritinse in Maize. Cornell Univ. Agric. Exp.
Sta. Memoir, 13, pp. 68.
Sturtevant, E. L.
1899. ier of Corn. U. S. Dept. of Agric. Office of Exper. Sta-
s Bull. 57: 7-103.

HOOKE’S MICROGRAPHIA

PROFESSOR LORANDE LOSS WOODRUFF

YALE UNIVERSITY

Brouocican research in general during the latter part
of the seventeenth century begins to be permeated with
an attention to details and with an intensive critical
analysis which is conspicuous by its absence in practic-
ally all but the masterpieces of previous times. Nor is
the explanation far to seek. The improvement of simple
lenses and the invention of the compound microscope
provided a method of increasing the apparent size of `
things which, in addition to revealing a new world of
‘‘animaleules’’ beyond the range of unaided vision,
brought to the attention of students finer details of struc-
ture of the higher animals and plants. But, as Sachs
has emphasized, the use of magnifying glasses contrib-
uted an advantage of still another kind—it taught those
who employed them to see scientifically and exactly. In
equipping the eye with increased powers the attention was
concentrated on definite points and observation had to
be accompanied by conscious critical reflection in order
to make the object, which is observed in part only by the
microscope, clear to the mental eye in all the relations of
the parts to each other and to the whole. Therefore, in
marked contrast with the very slow progress in obtaining
a mental mastery over the macroscopic morphological
features of plants and animals is the work of the early
students with the microscope such as Hooke and Grew in
England, Malpighi in Italy, and Swammerdam and
Leeuwenhoek in Holland.

The earliest clear appreciation of the importance of
studying nature with instruments which increase the
powers of the senses in general and the vision in par-
ticular, is found in a remarkable book by a remarkable

248 THE AMERICAN NATURALIST [Vou. LIII

man—the ‘‘Micrographia’’ of Robert Hooke, pae
by the Royal Society of London in 1665 (cf. Fig. 1).
The point of view of the author is well illustrated in the
following extracts from the preface:

It is the great prerogative of Mankind above other Creatures, that we
are not only able to behold the works of Nature, or barely to sustein our
lives by them, but we have also the power of considering, comparing,
altering, assisting, and improving them to various uses. And as this is
the peculiar priviledge of humane Nature in general, so it is capable of
being so far advanced by the helps of Art, and Experience, as to make
some Men excel others in their Observations, and Deductions, almost as
much as they do Beasts. By the addition of such artificial Instruments

MICROGRAPHIA:
OR SOME
Phyfological Defiřiptions
MINUTE BODIES
UAGNIREISG- GLASK S.

WITH
Osssavarions and Inquiries thereupon,

By R: HOO KE, Fellow of the Rovat Soctery.

oculo quantum contendere
im sag e Horat. Ep. lib. t.

LONDON „Printed by Jo. Martyn, and 7a. Allefiry, Printers tothe
YAL seku of andare vo be fold ar their Shop atthe Bein

Fia. 1.

and methods, there may be, in some manner, a reparation made for the
mischiefs, and imperfection, mankind has drawn upon it self, by negli-
gence, and intemperance, and a wilful and superstitious deserting the
Prescripts and Rules of Nature, whereby every man, both from a
deriv’d corruption, innate and born with him, and from his breeding
and converse with men, is very subject to slip into all sorts of errors.

. . . . +: . . . . . . . .

No. 626] HOOKE’S MICROGRAPHIA 249

The next care to be taken, in respect of the Senses, is a supplying of
their infirmities with Instruments, and, as it were, the adding of artificial
Organs to the natural; this in one of them has been of late years accom-
plisht with prodigious benefit to all sorts of useful knowledge, by the
invention of Optical Glasses. By the means of Telescopes, there is
nothing so far distant but may be represented to our view; and by the
help of Microscopes, there is nothing so small, as to escape our inquiry;
hence there is a new visible World discovered to the understanding. By
this means the Heavens are open’d, and a vast number of new Stars,
and new Motions, and new Productions appear in them, to which all the
antient Astronomers were utterly Strangers. By this the Earth it self,
which lyes so neer us, under our feet, shews quite a new thing to us, and
in every little particle of its matter, we now behold almost as great a
variety of eae as we were able parore to reckon up in the whole
Universe it s

It seems not Tanoba but that by these helps the subtilty of the
composition of Bodies, the structure of their parts, the various texture
of their matter, the instruments and manner of their inward motions,
and all the other possible appearances of things, may come to be more
fully discovered; all which the antient Peripateticks were content to

comprehend in two general and (unless further explain’d) useless words
of Matter and Form. From whence there may arise many admirable
advantages, towards the increase of the Operative, and the Mechanick
Knowledge, to which this Age seems so much inclined, because we may
perhaps be inabled to discern all the secret workings of Nature, almost

250 THE AMERICAN NATURALIST [ Vou. EHT

in the same manner as we do those that are the productions of Art, and
are manag’d by Wheels, and Engines, and Springs, that were devised
by humane Wit.

In this kind I here present to the World my imperfect Indeavours;
which though they shall prove no other way considerable, yet, I hope,
they may be in some measure useful to the main Design of a reformation
in Philosophy. ....

As for my part, I have obtained my end, if these my small Labours
shall be thought fit to take up some place in the large stock of natural
Observations, which so many hands are busie in providing. If I have
contributed the meanest foundations whereon others may raise nobler
Superstructures, I am abundantly satisfied; and all my ambition is, that
I may serve to the great Philosophers of this Age, as the makers and the
grinders of my Glasses did to me; that I may prepare and furnish them
with some Materials, which they may afterwards order and manage with
better skill, and to far greater advantage.

Toward the prosecution of this method in Physical Inquiries, I have
here and there gleaned up an handful of Observations, in the collection
of most of which I made use of Microscopes, and some other Glasses
and Instruments that improve the sense; which way I have herein taken,
not that there are not multitudes of useful and pleasant Observables, yet
uncollected, obvious enough without helps of Art, but only to promote
the use of Mechanical helps for the Senses, both in the surveying the
already visible World, and for the discovery of many others hitherto
unknown, and to make us, with the great Conqueror, to be affected that
we have not yet overcame one World when there are so many others to
be discovered, every considerable improvemnt of Telescopes or Micro-
scopes producing new Worlds and Terra-Ineognita’s to our view.

The author of this work was a versatile genius who
applied his powers to a wide field of endeavor—physics,
chemistry, mathematics, mechanics, architecture and
philosophy—fields which long since have expanded be-
yond the grasp of one man, and which, even in his own
time and by himself, might more profitably have been
coped with singly. It is impossible to adequately sur-
vey Hooke’s varied career within the limits imposed by
this paper, but the following extracts from his biography,
appended by Richard Waller to Hooke’s ‘‘ Posthumous
Works,’’ show what manner of man he was (ef. Fig. 5).

No. 626] HOOKE’S MICROGRAPHIA 251

Dr. Robert Hooke was Born at Freshwater, a Peninsula on the West
side of the Isle of Wight, on the eighteenth of July, being Saturday,
1635, at twelve a Clock at Noon, and Christened the twenty sixth follow-
ing by his own Father Minister of that Parish.

Schem:Xt.

Fig: i;

we

4 P
ee

From Westminster-School he went to the University of Oxford, in
1653. but as ’tis often the Fate of Persons great in Learning to be
small in other Circumstances, his were but mean. I find that he was a
Student of Christ-Church, tho’ not of the Foundation, but was, as I
have heard, a Servitor of one Mr. Goodman, and took his Degree of
Master of Arts several Years after, about 1662, or 1663.

About the Year 1655, he began to shew himself to the World, and
that he had not spent his Juvenile Years in vain; for there being a
Concourse at that time of extraordinary Persons at Oxford, each of
which afterwards were particularly distinguish’d for the great Light

252

THE AMERICAN NATURALIST

qii2ifks WHILE IHH ETN Maier
E ENRE ERP nadia
EREE 3 ETERNE TEE E FEE TE ĀZEŠ 4
Sa bse Pau ag laste
= ea begs £39 HIERNE SHH THS je
< FP Gah REH edie ite Hee
SZZERSSSSES LES LSS SESBEELES E Begses
AE RERA]
TRER HERUR HHE
eii
. x cf-F y 325 = $
ERE Enne
HEHU asia He Tea dee
me PEER Ears Hats SeED
menr qT qg LAU gimine pos
Fegi isg BINETEN EENE
HE Hiden Hing e
ana nen a ee pn
cine ann (iy rn Bis aa
e hids Ei is RELEE TER HIRR liga
o SERRI eiai gristi e iitgied eirig
o PERA geen EPEA 55 spicty FE S$ RETIE
-£ Fi; i 2 = S = A
ee r eters Crs E HAE
5 pedigy ity EEE BEE GEFs APARE
P RUHE HERIR AR Tea:
225225 $iSizeges Soest seers execs
BHR ma aine n
AERE HRR li Wigs
UR a SREI
SUL HE HEEE
HOU ae g mR
RUHEN RASHER rg eel
< ESeseiehaiaageceieessscsacsgeddy | & PRH
Sabie: | ip de
R EO HNT E T REH
: ATH ana a | 35 fp He
=z É HERA i preg tt i RR a3 LAS ie
O sacahes 238 ds Ge RH 5 a HT 25
e ae iaaa eii
5 ell aisn | 24 fii
=i S. ig 2 £ oy sz He > FP
Sle E ARE
HHH HKH HR all j lag

112

[Von. LIII

(Part 1.)

FIG. 4.

No. 626] HOOKE’S MICROGRAPIIIA

of thofe

breath, nor
one

herefore there

juices of Ve-

Vi

Orginiza-

lefs

divers Emi-

is common to all Vegetables, as may appear by fos

lity
A i À
ie cod ingesiocn ad East:
t k
ks which, haviog that liberty granted ray

that moft Iluftrious Society, I have hereunto ad joyn'd.

were made

ope,
pa

ifcover a

‘ough concrete,and that each Ca-
rate from any of the reft, without
ing films, fo that I could no more blow
ice,then I could through a piece of
A nor with
Gigs on OF
conclude, that t
fome of thele kind

l
to the contain’d fluid
om ed
uve has
digre amd ae ge ipo ‘tis not im
cr, if belp'd with better Avicro/copes,may
be fo, feems with i
: of fenleive
Obfervations that
and Dr. Clark,
Thereare four Plants, two of which are little fhrub Plants,
with q litle fhort ftock, about an Inch above the ground, from

not
le,that

MICROGRAPHIA

of folid ot hardned froth,or a cengeries of very {mall bubbles confolidated

ftiffas well as ti
ind
inde ok fubfta
of an Elder,
with p
', yet I cannot thence
ich the Sucews mutritine
them; for, in
juke ana by
green’ fay al
great
fuch thofe
jeer in
Epele
banomena
of the Royal Society on
account eed in pkita

116
vern,
hi 29
throu
Cork,
But
ca’
are
Rail ge
which
Now.
a
themfe!
not hitherto been able to
it feems very
of Animal
whereby to
but that fome
in time
from the
tion that is
inftrudt

on the Humble and Senfible Plants in M" Chiffin's

or
pund
byfician, Dottor Clar.

iver

this may
Pi
‘den in Saint James's Park, made Auguft the 9™ 1661,

Bubble, or
ind of hole in the
qi the |
ive then curious
Mr, Evelin, Dr. Henfhaw,

t Members

ort
any other way I have
nent,

of ah

lent ri

in that form, into
ti deteĝ.
And that

Ovfervatons
Gar.
Prefer,

it
at all
h
mar dis eof, fer,
arbor minio pila videatur. Which
e tree, fubftance, and manner of grow-
> aa onion ae
I bave found that the say iak
Or pi hol
poy Sak ay xt roy ase dake
pane enag bog have much fuch a kind of

MICROGRAPHIA:

o
barkof
it,
erare
bark underneath remaining
d the Tree, and
that with other Trees)
fie, dor f procs, Lape
procera, Li
tice detralo juvatur,
é l, mira sterum
Cortex iz
qui r y
ita
confider'd, and the tree,
¢xamin'd, would, lam very
the origination of
Texture

bye iad
HUD HE EE
lle aai

i
Suge
ef

EERTE
$ ebay! EEE

= D
an

i i
s] pois
E

i

E
nse

oak

ight, and

fmooth,

whence are fpread feveral fticky branches, round, ftrei

ah
i Hil
TEs, a T si
pueh

(PART 2.)

Fica. 4.

253

254 THE AMERICAN NATURALIST [ Vou. LIII

they gave the Learned World by their justly admired Labours; he was
soon taken notice of, and for his Facility in Mechanick Inventions much
priz’d by them

= The same Year I contriv’d and made many trials about the Art of
flying in the Air, and moving very swift on the Land and Water, of
which I shew’d several Designs to Dr. Wilkins then Warden of Wadham
College, and at the same time made a Module, which, by the help of
prings and Wings, rais’d and sustain’d itself in the Air; but fin
ing by my own trials, and afterwards by Sane that the esis
of a Mans Body were not sufficient to do anything considerable o
that kind, I apply’d my Mind to contrive a way to make artificial
Muscles; divers designs whereof I shew’d also at the same time to Dr
Wilkins, but was in many of my Trials frustrated of my expectations.”?
What is mentioned here of his attempts about flying, is confirm’d by
several Draughts and Schemes upon Paper, of the Methods that might
be attempted for that purpose, and of some contrivances for fastening
suecedaneous Wings, not unlike those of Bats, to the Arms and Legs of
a Man, as likewise of a Contrivance to raise him up by means of Hori-
zontal Vanes plae’d a little aslope to the Wind, which being blown
round, turn’d an endless Serew in the Center, which help’d to move the
Wings, to be manag’d by the Person by this means rais’d aloft... .

Soon after the beginning of the Royat Socriety, viz. about April
1661. a Debate arose in the Society, oceasion’d by a small Tract Printed
in 1660. about the cause of the rising of Water in slender Glass Pipes,
higher than in larger, and that in a certain proportion to their Bores;
this Discourse was wrote and Publish’d by Hooke; the Explication of
which diffieult Phenomenon made him the more regarded. The sum of
his Reasonings upon this Subject he Publish’d afterward, Micrography
Observ. the 6th. in which there are several very curious and then new
Remarks and Hints; as to the Nature of Fluidity and Gravity, which
last is farther prosecuted in his Treatise of Springs, with other excel-
lent Subjects, to which the Inquisitive are referr’d for a more ample
satisfaction.

This, together with his former Performances, made him much re-
spected by the R. Society, and on the fifth of November 1662. “ Sir
Robert Moray propos’d a Person that was willing to be entertain’d as a
Curator by the Society, offering to furnish them every day when! they
met, with three or four considerable Experiments; which Proposition
was unanimously receiv’d, Mr. Hooke being nam’d to be the Person; and
accordingly the next Day of their meeting on the twelfth of November
he was unanimously accepted and taken as Curator, with the Thanks
of the Society order’d to Mr. Boyle for dispensing with him for their

1 From Hooke’s diary.

No. 626] HOOKE’S MICROGRAPHIA 255

use, and order’d that Mr. Hooke should come and sit among them, and
both bring in every Day three or four of his own Experiments, and
take care of such others as should be recommended to him by the
Society.”

wW ORK KS |

ROBERT HOOKE M.D. S.RS.

Ta ra Tan

DISCOUR SES,

of the Tut:

ROYAL SOCIETY.

I. The prefent Deficiency of pppoe Foiss orxy is difcourfed of,
with the Methods of rendering it more certain and hase prg
nd Eksa Li red of, particularly

‘ela
IIL An Hypothetical ak Mio dit WY the paame

I
of by the Mind in its tion ma) be coat Onan
IV. An Hypothesis ication Of the caufe RANITY, Or GRA-
VITATION, Magnetism,
Y, Dicsures of Ean ruana e s dheir Cis and Efeis, and Hiftories
of feveral ; to which are annext, -Ph ui eon ication ‘of feveral of the Fa-

‘bles in Ove 's Metamorphofes, yery.d t from other Mythologick Inter-
vE Lettures for for improv: om and AstTRONOMY, with
the Defcriptions of rma. Pp new pee ‘et ful cries ota Fea Contrivance!
the whole full of curious Difquifitions and

Tothefe Discourses is prefix the AuT Ho R’s Lire, giving an Ac-

= "o Da Studies and Employments, with an a se man nF apa
panera trivances and Inventions, by hii

Cars ete a: rrn to the Royal Society.

BLISH’D

By RIC HARD WALLER, R.S. Secr.

bo a:
Printed by Sam Smatitand Banh Ware a ages the
Royal Society) at the Premces Arms in St. Paai’s C frke bih ard.

ee E

a5 3 Ci tl = — dedd Ric: Daler Rey Juut kpda
` Fra. 5.

From this time the Societies Journals gave sufficient: Testimonials of
his Performances, all which' woul too many to particularize here,
therefore I shall only touch upon some of the chief.

At several Meetings of the Society in 1663, and 4. he produe’d his
Microscopical Observations, and read the Explications and Discourses
made upon them, which were®after publish’d in his Micrographia, at the
beginning of the year 1665.

p Sir John Cutler having founded a Lecture, and settl’d an Annual
Stipend upon Robert Hooke, M.A. of fifty Pounds during Life (entrust-
ing the President, Council and Fellows of the said Society to direct and
appoint the said Mr. Hooke as to the Subject and Number of his Lee-

256 THE AMERICAN N4TURALIST [Vou. LIII

tures) the Society order’d several of their Members to wait upon Sir
John Cutler, with their Thanks for his particular Favour to a worthy
Member, and for that — a Confidence he hath hereby exprest
towards their whole Body, ete.”

Fi ihis time he EERS in almost at every Meeting Experiments,
Observations, Schemes of new Instruments and Inventions, or something
LUP BB to the advancement of Knowledge, and very frequently read

utlerian Lectures, of many whereof he publish’d, the most material
a in his Tracts Printed at different times, in Quarto, call’d Lectures
and Collections, &e. comprizing compendiously in one continu’d Dis-
course, the chief Matters and Subjects handled in several Lectures.

Thus the generous Ardor with which the Royau Society was inspir’d
continued ’till the Year 1665, when, by reason of the great Mortaity
then reigning, they were oblig’d to ani and break up their Weekly
Meetings till the fourteenth of March 1663. .

The dreadful Conflagration of a great part of the City of London

House in the Strand, where, by the favour of the, then Duke of Norfolk,
they prosecuted their former Inquiries, their first Meeting at Arundel
House being on the ninth of Jan. 166 .

n the nineteenth of Sep. 1666. he produe’d a Module he had
design’d for the Rebuilding of the City, with which the Society were
very well pleas’d, and Sir John Laurence the then Late Lord Major, ad-
dress’d himself to the Society, expressing the present Lord Majors and
Aldermens liking thereof, as also their desire that it might be shewn to
his Majesty, they preferring it far before the Model drawn up by the
City Surveyor.

What this Model was, I cannot so well determine, but I have heard
that it was design’d in it to have all the chief Streets as from Leaden-
Hall corner to ies ting and the like, to lie in an exact strait Line, and
all the other cross Streets turning out of them at right Angles, all the
Churches, publick Buildings, Market-places, and the like, in proper and
convenient places, which, no doubt, would have added much to the
Beauty and Symmetry of the whole. How this came not to be accepted
of I know not, but it is probable this might contribute not a little to his
being taken notice off by the Magistrates of the City, and soon after
made Surveyo

The seas of the City, according to the Act of Parliament, re-
quiring an able Person to set out the Ground to the several Proprietors,

2From the Journal of the Secretary of the Royal Society, November 9,

664.

*

.

No. 626] HOOKE’S MICROGRAPHIA 257

Mr. Hooke was pitch’d upon, and appointed City-Surveyor for that
difficult Work, which being very great, took up a large proportion of
his Time, to the no small hindrance of his Philosophical Disquisitions.

In this Employment he got the most part of that Estate he died pos-
sessed of, as was evident by a large Iron Chest of Money found after his
Death, which had been lock’d down with the Key in it, with a date of the
Time, by which it appear’d to have been so shut up for above thirty
Years: In this was contain’d the greatest part of what he left behind
him, which was to the value of many thousands in Gold and Silver.
That he might by this place justly acquire considerable Estate, I think
cannot be en ag,

Mr. ‘Oldenburgh, the sbi Saidiy; dying in the time of the Societies
Recess, 1677. Mr. Hooke was desir’d to take his place, and take the
Minutes of what considerable Matters past, which he did on the twenty
fifth of October 1677. and the same day produe’d his Waterpoise and
shew’d the nicety thereof.

From that time he officiated in that Place, as well as his Curatorship,
shewing several Experiments and Instruments in order to explain the
Gravitation and Alterations in the Air by Vapours, ete. Contriving an
Air-poise to shew the different specifick Gravity of the Air by a large
thin ball of Glass Feryat areir

Pioi this ENR he ae Nise odsbptial Chiiecoinic on Atlee
in Peper-water, and other Seeds steeped in Water, confirming Monsieur
Leuenhook’s Assertions, and propos’d some Improvements of Micro-
scopes.

Apr. 25. 1678. he shew’d an Experiment farther to explain the action
of a Muscle, “ which was by a Chain of small Bladders fastened together,

“so as by blowing into one Pipe, the whole might be atthe fill’d,

and by that means contracted, supposing the Fibres of the Muscles
which seem’d like a Necklace of Pearl in the paced ee be
fill’d with a very agill Matter, which he thought most likely to be Air,
which being included in so thin Skins, was easily wrought upon by
Heat, Cold, or the acting Properties of the Liquors that pass between
them, and so perform the lengthening and contracting of the Muscles.

Aug. 1678. he read several Discourses, and shew’d Experiments in
order to confirm his eis of a and springy Bodies. .. .

Thus I tees ad some of his Pees A ae “It must be
confessed that the latter part of his Life was nothing near so fruitful
of Inventions as the former; tho’ it is certain he had a design to repeat
the most part of his Experiments, and finish the Accounts, Observations

258 THE AMERICAN NATURALIST [ Vou. LIII

and Deductions from them, and had an Order for the Societies bearing
the Charge thereof, in June 1696. when he propos’d likewise to perfect
the Description of all the Instruments he had at any time contriv’d;

but by reason of his increasing Weakness and a general Decay, he was
absolutely unable to perform it, had he desir’d it never so much.

Thus he liv’d a dying Life for a considerable time, being more than a
Year very infirm, and such as might be eall’d Bed-rid for the greatest
part, tho’ indeed he seldom all the time went to Bed. . . . being emaci-
ated to the utmost, his Strength wholly worn out, he dy’d on the third
of March 1703. being 67 Years, 7 Months, and 13 Days Old.

His Corps was decently and handsomely interr’d in the Church of St.
Hellen in London, all ‘he Members of the Royau Socrery then in Town
attending his Body to the Grave, paying the Respect due to his extraor-
dinary Merit.

As to his Person he was but despicable, being very crooked, tho’ I
have heard from himself, and others, that he was strait till about 16
Years of Age when he first grew awry, by frequent practicing with a
Turn-Lath, and the like incurvating Exercises, being but of a thin weak
habit of Body, which inereas’d as he grew older, so as to be very re-
markable at last: This made him but low of Stature. ... He went
stooping and very fast (till his weakness a few Years before his Death
hindred him) having but a light Body to earry, and a great deal of
Spirits and Activity, especially in his Youth.

He was of an active, restless, indefatigable Genius even almost to the
last, and always slept little to his Death, seldom going to Sleep till
two three, or four a Clock in the Morning, and seldomer to Bed, often
continuing his Studies all Night, and taking a short Nap in the Day.
His Temper was Melancholy, Mistrustful and Jealous, which more in-
ereas’d upon him with his Years. He was in the beginning of his being

made known to the Learned, very communicative of his Philosophical
Diseoveries and Inventions, till some Accidents made him to a Crime
close and reserv’d. He laid the cause upon some Persons, challenging

his Discoveries for their own, taking occasion from his Hints to perfect
what he had not; which made him say he would suggest nothing till he
had time to perfect it himself, which has been the Reason that many
things are lost, which he affirm’d he knew. He had a piercing Judg-
ment into the Dispositions of others, and would sometimes give shrewd
Guesses and smart Characters.

It must be confess’d that very many of his Inventions were never
brought to the perfection they were capable of, nor put in practice till
some other Person either Foreigner or of our own Nation cultivated
the Invention, which, when Hooke found, it put him upon the finishing
that which otherwise possibly might have lain ’till this time in its first

.

No. 626] HOOKE’S MICROGRAPHIA 259

Defects: Whether this mistake arose from the multiplicity of his Busi-
ness which did not allow him a sufficient time, or from the fertility of
his Invention which hurry’d him on, in the quest of new Entertain-
ments, neglecting the former Discoveries when he was once satisfied of
the feazableness and certainty of them, tho’ there wanted some small
matter to render their use more practicable and general, I know not... .

Whatever the answer may be, Hooke’s first and best ~
known work, the Micrographia, at once epitomizes the
versatility of his genius as well as his apparent inability
to see one problem through to a finish. To quote from a
review of the work in the Philosophical Transactions,
No. 2, Monday, April 3, 1665:

The Ingenious and knowing Author of this Treatise, Mr. Robert
Hook, considering with himself, of what importance a faithful History

ry
what portion he can, hath lately published a Specimen of his abilities
in this kind of study, which certainly is very welcome to the Learned
and Inquisitive world, both for the New discoveries in Nature, and the
New Inventions of Art.

To this end, he hath made a very curious Survey of all kinds of
bodies, beginning with the Point of a Needle, and proceeding to the
Microscopical view of the Edges of Rasors, Fine Lawn, Tabby, Watered
Silks, Glass-canes, Glass-drops, Fiery Sparks, Fantastical Colours,
Metaline Colours, the Figures of Sand, Gravel in Urine, Diamonds in
Flints, Frozen Figures, the Kettering Stone, Charcoal, Wood and other
Bodies petrified, the Pores of Cork, and of other substances, Vegetables
growing on blighted Leaves, Blew mould and Mushroms, Sponges, and
other Fibrous Bodies, Sea-weed, the Surfaces of some Leaves, the sting-
ing points of a Nettle, Cowage, the Beard of a wild Oate, the seed of the
Corn-violet, as also of Tyme, Poppy and Purslane. He continues to
describe Hair, the scales of a Soal, the sting of a Bee, Feathers in gen-
eral, and in particular those of Peacocks; the feet of Flies; & other In-
sects; the Wings and Head of a Fly; the Teeth of a Snail; the Eggs of
Silk-worms; the Blue Fly; a water Insect; the Tufted Gnat; a White
Moth; the Shepherds-spider; the Hunting Spider, the Ant; the wan-
ding. Mite; the Crab-like insect, the Book-worm, the Flea, the Louse,
Mites, Vine-mites. He concludeth with taking occasion to discourse of
two or three very considerable subjects, viz. - ve of the Rays
of Suis in ale Air; the Fixt starrs; the Moo

260 THE AMERICAN NATURALIST [ Vou. LIII

In representing these particulars to the Readers view, the Author
hath not only given proof of his singular skil in delineating all sorts of
Bodies (he having drawn all the Schemes of these 60 Microscopical

very noble contemplations. Here are found inquiries concerning the
Propagation of Light through differing mediums; concerning Gravity;
concerning the Roundness of Fruits, stones, and divers artificial bodies;
concerning Springiness and Tenacity; concerning the Original of Foun-
tains; concerning the dissolution of Bodies into Liquors; concerning
Filtration, and the ascent of Juices in Vegetables, and the use of their
Pores. Here an attempt is made of solving the strange Phenomena of
Glass-drops; experiments are alleged to prove the Expansion of Glass
by heat, and the Contraction of heated-Glass upon cooling; Des Cartes
his Hypothesis of colours is examined: the cause of Colours, most
likely to the Author, is explained: Reasons are produced, that Reflection
is not necessary to produce colours, nor a double refraction: some cou-
siderable Hypotheses are offered, for the explication of Light by Motion;
for the producing of all colors by Refraction; for reducing all sorts of
colors to two only, Yellow and Blew; for making the Air, a dissolvent
of all Combustible Bodies: and for the explicating of all the regular
figures of Salt, where he alleges many notable instances of the Mathe-
maticks of Nature, as having even in those things which we account
vile, rude and coorse, shewed abundance of curiosity and excellent
Geometry and Mechanism. And here he opens a large field for in-
quiries, and proposeth Models for prosecuting them

He goes on to offer his thoughts about the Pores of iodine: and a kind
of Valves in wood ; _ t spontaneous generation arising from the

of the

found in the leaf of a Nettle, how the stinging pain is created, and
i e sub-

contrivance and fabrick of Feathers for Flying. He delivers those par-
ticulars about the Figure, parts and use of the head, feet, and wings of a

y, that are not common. He observes the various wayes of the gen-
erations of Insects, and discourses handsomely of the means, by which

hunting Spider, and other Spiders and their Webs. And what he notes
of a Flea, Louse, Mites and Vinegar-worms, cannot but exceedingly
please the curious Reader.

No. 626] HOOKE’S MICROGRAPHIA 261

Having dispatched these Matters, the Author offers his Thoughts for
the explicating of many Phenomena of the Air, from the Inflexion, or
from a Multiplicate Refraction of the rays of Light within the Body of
the Atmosphere, and not from a Refraction caused by any terminating
superficies of the Air above, nor from any such exactly defin’d super-
ficies within the body of the Atmosphere... .

e coneludeth with two Celestial Observations; whereof the one im-
ports, what multitudes of Stars are discoverable by the Telescope, and
the variety of their magnitudes . . . the other affords a description of a
Vale in the Moon, compared with that of Hevelius and Ricciolo; where
the Reader will find several curious and pleasant Annotations . .
about the variations in the Moon, and its gravitating principle, Kigeuaat
with the use, that may be made of this Instance of a gravity in the
Moon.

As to the Inventions of Art, described in this Book, the curious
Reader will there find these following:

1. A Baroscope, or an Instrument to shew all the Minute Variations
in the Pressure of the Air; by which he affirms, that he finds, that before
and during the time of rainy weather, the Pressure of the Air is less, .

Hygroscope, or an Instrument, whereby the Watery stidi,
oitis in the Air, are discerned, which the Nose it self is not able to
find. Which is by him full described in the Observations touching the
Beard of a wild Oate, by the means whereof this Instrument is con-
trived.

3. An Instrument for graduating Thermometers, to make them Stand-
ards of Heat and Cold.

4. A New Engin for Grinding Optik Glasses, by means of which he
hopes, that any Spherical Glasses, of what length soever, may be
speedily made... .

5. A New Instrument, by which the Refraction of all kinds of Liquors
may be exactly measured, thereby to give the Curious an opportunity of
making Trials of that kind, to establish the Laws of Refraction. .. .

Lastly, this Author despairs not that there may be found many Me-
chanical Inventions, to improve our Senses of Hearing, Smelling, Tast-
ing, Touching, as well as we have improved that of Seeing by Optick
Glasses.

Thus the ‘‘Micrographia’’ is obviously something
more than ‘‘Some Physiological Descriptions of Minute
Bodies made by Magnifying Glasses’’—it is a demon-
stration of the advantages to be gained by the use of ar-
tificial devices of precision in studying nature. The
book is replete with singular anticipations of later dis-
coveries and inventions by other workers and ‘‘it will
hardly be deny’d that there are more excellent Philo-

262 THE AMERICAN NATURALIST [ Vou. LIII

sophical Discoveries and Hints, than in most extant of
its bulk.” It contains the first study of the ‘‘ fantastical
colours’? of thin plates with a partial explanation by in-
terference; a theory of light as a ‘‘very short vibrating
motion’’ transverse to straight lines of propagation
through a ‘‘homogenous medium’’ (p. 56). Heat is
stated to be ‘‘a property of a body arising from the
motion or agitation of its parts” (p. 37); Fluidity is ‘‘but
an effect of a very strong and quick shaking motion,
whereby the parts are, as it were, loosened from each
other, and consequently leave an interjacent space or
vacuity” (p. 41); while ideas in regard to combustion
are clearly outlined (p. 103) which foreshadow those
reached by Mayow.

But the biologist’s interest in the ‘‘Micrographia’’ is
chiefly in Hooke’s application of his improved compound
microscope (Fig. 2) to the study of animals and plants.
At this time Malpighi, Grew, Leeuwenhoek and Swam-
merdam were engaged in studies, with simple lenses or
compound microscopes, on the secrets of the finer struc-
ture of organisms which were to give them higher rank
in biological history than Hooke’s desultory work in this
field. Hooke, as has been said, was interested primarily
in demonstrating the usefulness of his microscope and
his belief that in inventions for the ‘‘improvement of
the senses’’ lay the key to a more profound understand-
ing of nature. This he accomplished and therefore, en-
tirely aside from the other remarkable qualities of the
‘‘Micrographia,’’ the book holds a unique place in the
history of biology. It paved the way, as it were, for the
more special, profound and methodical studies of the
contemporary founders of the morphology of organisms
by creating a considerable interest in microscopy, and in
addition proved to be for over a century the standard
source from which writers on the microscope gleaned
much information and many figures.’

3 E. g., L. Joblot, 2d Ed., 1754; H. Baker, 1742; M. F. Ledermiiller, 1760-
1763; etc. In 1745 Baker reprinted and explained the plates of the Micro-
graphia.

No. 626] HOOKE’S MICROGRAPHIA 263

Among the large variety of observations made by
Hooke, which are cited in Oldenburg’s review’just quoted
in extenso, one in particular claims the attention of the
biologist. This is ‘‘Observation XVIII. Of the Schema-
tisme or Texture of Cork, and of the Cells and Pores
of some other such frothy Bodies.’’ Here are clearly de-
scribed for the first time the ‘‘little boxes or cells”? of or-
ganic structure, and his use of the word ‘‘cell’’ is re-
sponsible for its application to the protoplasmic units of
modern biology. This observation, together with the
plate, is presented in facsimile in Figs. 3 and 4. In
Hooke’s treatise on ‘‘ The method of Improving Natural
Philosophy,’’ included in the volume of his posthumous
works (p. 28), this observation on cells is selected by
Hooke to illustrate his method of scientific inquiry.

Again, ‘‘Observation XXV. Of the stinging points
and juice of Nettles, and some other Venomous Plants”?
is accompanied by a figure of the lower side of a nettle
leaf in which the outlines of the epidermal cells are well
delineated and, as Miall remarks, ‘‘there is something
very like a nucleus in one of them, but this may be acci-
dental.’’ However, Hooke did not recognize any rela-
tionship between the structures he observed in the nettle
and in the cork.

As an appendix to his gbservations on cork, the author
relates some experiments on Mimosa in which he attrib-
utes the ‘‘motion of this Plant upon touching ... to a
constant intercourse betwixt every part of this Plant and
its root, either by a circulation of its liquor, or a constant
pressing of the subtiler parts of it to every extremity of
the Plant’’—a partial anticipation of the modern idea
of turgescence (cf. Figs. 3 and 4). The ‘‘Observation on
Petrify’d wood and other Petrify’d bodies”’ is interest-
ing because the author takes quite a modern point of
view in regard to fossils (cf. Fig 4, Part I).

And so we might continue—but as the reviewer re-
marks in the Journal des Scavans, December, 1666: ‘‘ This
Book contains more than can be taken notice of in an

264 THE AMERICAN NATURALIST [ Vou. LIII

Extract,’’ and we conclude the survey of this man of the
past still using the words of the past:

All his Errors and Blemishes, were more than made amends for, by
the Greatness and Extent of his natural and acquired Parts, and more
than common, if not wonderful Sagacity, in diving into the most hidden

Changes, to her last and utmost Recesses; so that what Ovid said of
Pythagoras may not unfittly be apply’d to him.

Mente Deos adiit, et quae Natura negavit
Visibus humanis, oculis ea Pectoris hausit.

There needs no other Proof for this than the great number of Experi-
ments he made, with the Contrivances for them, amounting to some
hundreds; his new and useful Instruments and Inventions, which were
numerous, his admirable Facility and Clearness, in explaining the Phe-
nomena of Nature, and demonstrating his Assertions; his happy Talent
in adapting Theories to the Phenomena observ’d, and contriving easy
and plain, not pompous and amusing Experiments to back and prove

eories; proceeding from Observations to Theories, and from
Theories to farther trials, which he often asserted to be the most proper
method to succeed in the interpretation of Nature. For these, his happy
Qualifications, he was much respected by the most learned Philosophers
both at home

e
from them, possibly, he might have stood in the Front. But humanum
est errare.

4 Waller, op. cit.

SHORTER ARTICLES AND DISCUSSION.

SIAMESE, AN ALBINISTIC COLOR VARIATION IN
CATS

COMPARATIVE studies of color inheritance in mammals have
shown that pigment production throughout the group is due to
similar processes and to genes probably homologous. These
studies have shown, for example, that the pink-eyed albino con-
dition seen in white rabbits, white rats, white mice and white
guinea-pigs behaves in all cases as a simple recessive in crosses.
It is probably due to variation in the same (t. e., in an homo-
logous) genetic locus in all these rodents. In its usual form
albinism consists in a complete absence of pigmentation from
the ectoderm of the embryo and from all derivatives of that
germ-layer in the adult animal. This includes, not only the
hair, but also the retina and iris of the eye. Such is the condi-
tion seen in the white mouse, the white rat, and the ‘‘Polish’’

r ‘‘Russian’’ rabbit. But this same locus may apparently
undergo a different change which, while it behaves as the per-
fect allelomorph of the pure white albino variation, differs from
it in that it allows a certain amount of pigment to be produced,
more particularly in the retina of the eye and in the hair at
the extremities of the body (nose, ears, tail and feet). At times
a small amount of pigment is formed elsewhere throughout the
coat. This condition is best known in the ‘‘Himalayan’’ rabbit.
Clear white albinism of the Polish rabbit is an allelomorph of
Himalayan albinism. In the guinea-pig only the Himalyan
type of albinism is known; in rats and mice, only the Polish
type is known.

In the guinea-pig, Wright has demonstrated the existence of
two other albino allelomorphs, which apparently are distinct
mutations of the same genetic locus. These are found in the red-
eyed and in the dilute varieties described by him. Among rats
Whiting and King have demonstrated the existence of a variety
comparable with the dilute varieties of guinea-pigs and which
they call ‘‘ruby-eyed.’’ It behaves as an allelomorph of ordi-
nary albinism in crosses.
` White spotting of colored animals, sometimes called ‘‘ partial
albinism,’’ is an entirely different variation, due to variation in
a different locus. True albinism and spotting may by suitable
crosses be made to coexist in the same individual. In this way

265

266 THE AMERICAN NATURALIST [Vou. LII

I have frequently produced spotted Himalayan rabbits, which
would show particular types of white spotting, as Dutch or
English, on the feebly pigmented Himalayan background (as
has also Punnett), and Wright has produced whole series of
varieties of spotted red-eyed and spotted dilute guinea-pigs.

Among certain rodents pink-eyed varieties occur which are
due to variation in a genetic locus wholly distinct from that
which is responsible for albinism. Such are the well-known
pink-eyed varieties of mice having colored coats. Here the
retina and the fur alike have a greatly reduced amount of black
and brown pigmentation as compared with normal individuals,
though yellow is unaffected. Pink-eyed rats and pink-eyed
guinea-pigs are similar in appearance and in genetic behavior
to pink-eyed mice. When crossed with the albino variety of
the same species, they produce fully colored offspring as regards
both eye and coat. The gene for pink-eye is thus seen to be
complementary to the gene for albinism, with which it is known
to be ‘‘linked’’ in rats and mice. Whether the two are also
‘‘linked’’ in guinea-pigs has not yet been ascertained.

Among mammals other than rodents albino and pink-eyed
varieties are not certainly known to occur, though white-spotted
and black-eyed white varieties are common. It is-thus an open
question whether the same genetic loci are found among them
as among rodents. Bateson has pointed out similarities be-
tween a color variety of cat, the so-called Siamese, and the
Himalayan variety of rabbit. Both are born white or nearly
white and later become more heavily pigmented. I may add
(2) that both are inherited as recessives and (3) that in both
varieties yellow pigment is largely or wholly suppressed, which
is characteristic r op albino variation, but not of the pink-eye
variation of roden

Wright has ii that blondism among human beings
(which when extreme in character is commonly known as al-.
binism) is similar in nature to the albinism of rodents, being a `
graded series of allelomorphs similar to the series which he has
described in the guinea-pig.

It thus appears probable that the same genetic locus, which
occurs in rodents and which has been called the ‘‘color factor,”
occurs also in other mammals, including man.

The case of the Siamese cat has seemed to me for some years
deserving of more careful study. Lacking opportunity for such
study myself, I sent out an inquiry several years ago through

No. 626] SHORTER ARTICLES AND DISCUSSION 267

the pet-stock journals for information about Siamese cat crosses

A single reply has just come to hand, but from an authoritative
source. A doctor, who prefers to remain anonymous, resident
in an extensive institution in England and a fancier of Siamese
cats, has employed his leisure, and the unusual opportunities
afforded by his position, in studying the genetic behavior of
Siamese cats in crosses with other varieties. He regards as
characteristics of the Siamese breed a peculiar quality of voice
and which characters often are seen in first
generation crosses and so would seem to be inclined to domi-
nance. But the distinetive Siamese eolor, he states, is never
seen in F, individuals, ‘‘although quite a number show a mid-
way color. At a glance you would say they were black, but on

more careful examination you see they are near the color of the
Siamese ears, seal brown. Most first crosses in my experience
are black or seal, but some tortoise shell, or tortoise shell and
white, or black and white.’’ These statements indicate the usual.
behavior of yellow and of white-spotting in cat crosses. (See
Whiting, 1918.) The Siamese color is evidently an independ-
. ent character incompletely recessive in F,. The doctor contin-
ues his account with a brief statement concerning a back cross
of F, with pure Siamese. ‘‘I have a first cross female, black
seal color, marked cross eyes, Siamese voice. She has been twice
mated with a pure Siamese male. In her first litter she had two
pure Siamese, perfect Siamese color. Unfortunately both died
of distemper when about three months old. Her second mating
resulted in one pure Siamese which is still alive. It is about
five months old and is perfect in all Siamese points and fit to
win [at shows]. Presumably the same sort of back-cross mat-
ings as these would produce also kittens similar to the F, mother
in color character, although no mention is made of them in these
notes. The information given suffices to show the segregation
of Siamese color as a recessive character in generations later
than F,. The doctor confirms the observation of others as to
the deficient pigmentation of the eye, a point of resemblance
with allelomorphs of true albinism, as seen, for example, in red-
eyed guinea-pigs (Castle and Wright), and in ruby-eyed rats
(Whiting). He says: ‘‘The reflex which the Siamese cat shows
in the dark is worth notice. It looks blood red and must be due
to absence of pigment in the retina.’’ A further point of re-
semblance with albinism is its distinctness from dilution as seen
in ‘‘blue’’ varieties. The doctor speaks of having produced

268 THE AMERICAN NATURALIST [Von. LII

four Siamese which are ‘‘blue-pointed,’’ presumably as a re-
sult of crosses with maltese, which are blue pigmented. An
exactly similar combination I have recently secured in crossing
rabbits, obtaining Himalayans with blue points in F, from a
cross between ordinary black-pointed Himalayans and a self-
colored rabbit which carried blue as a recessive character.

To summarize, we have the following indications that Siamese
coloration in cats is a form of true albinism similar to that of
the Himalayan rabbit, and still more closely resembling the
ruby-eyed rat and the red-eyed guinea-pig, all of which species
possess also more typical forms of albinism, but which are allelo-
morphs of those mentioned.

(1) Siamese coloration in cats is attended by a deficiency in
amount of pigmentation in both coat and eye. (2) Yellow pig-
ment is more affected than black or brown pigment. (3) The
pigmentation is less at birth than at a later period. (4) The
character is recessive in heredity. (5) It is distinct from
‘*blue’’ dilution since it can be combined with it by suitable
crosses.

Siamese in cats as far as reported occurs only in a non-agouti
form, as does Himalayan in rabbits bred for exhibition. But
by a eross with agouti rabbits, Himalayan rabbits are obtained
in F, which have agouti points. As this makes the contrast of
points with body less strong, fanciers’ standards do not recog-
nize the combination. Nevertheless the experiment shows agouti
to be due to a genetic factor distinct from Himalayan. If Sia-
mese in cats is also distinct from agouti, it may be expected that
a cross of Siamese with tabby would produce Siamese tabbies in
F., though the combination would probably not be pleasing to
the fancier.

W. E. CASTLE.
BUSSEY INSTITUTION.
BIBLIOGRAPHY
Bateson, W.
1913. Mendel’s agea of Heredity.
Castle, W. E., and S. W
1916. Studies of p in Guinea-pigs and Rats. Carnegie Insti-
tution of Washington, Publication No. 241.
Punnett, R. C.
1912. Inheritance of Coat-eolor in Rabbits. Journal of Genetics, 2.
Whiting, P., W
1918. Inheritance of Coat-color in Cats. Journ. Exp. Zool., 25, p. 539.
Whiting, P. W., and Helen Dean Kin,
1918. Biti Dilute Gray, a Third n in the Albino Series
of the Rat. Jour. Exp. Zool., 26, p. 5

No. 626] SHORTER ARTICLES AND DISCUSSION 269

THE MORPHOLOGICAL BASIS OF SOME EXPERI-
MENTAL WORK WITH MAIZE

OF all the plants that have been made to contribute to our
knowledge of the principles of evolution and heredity in the last
twenty years, probably none holds a more conspicuous place than
Indian corn. The technique of its manipulation is compara-
tively simple, and it exhibits an extreme variability, which is
almost unique in extending to the endosperm; the behavior of a
large number of its characteristics has been found amenable to
a Mendelian interpretation and has aided materially in estab-
lishing present-day views of heredity. Indeed, maize shares
with Pisum the distinction of having been the means of the estab-
lishing of Mendelism itself, for it was in connection with their
work on maize that Correns and De Vries discovered Mendel’s
paper. Since then its genetic behavior has been studied in detail
by a number of investigators, and there is probably no other one
plant that furnishes such a wealth of material illustrative of the
principles of heredity.

The writer has in recent years had the opportunity of examin-
ing in more or less detail this same plant from the morphological
point of view, and it has been found that we are far more famil-
iar with the Mendelian behavior of some of its characteristics
than we are with the characteristics themselves. This has led to
some results illustrative of the need of very close coordination
between genetics and morphology.

In one of the numerous experiments made by East and Hayes,
an attempt was made to interpret the Mendelian behavior of
the irregularity of the rows of grains on the ear of corn. The
ratios produced in the breeding experiments' were not very sig-
nificant, and, after suggesting the possibility of ‘‘monohybridism
with reversed dominance,’’ ‘‘fluctuating dominanee,’’ ete., they
finally conclude that ‘‘it seems probable that a more complex set
of conditions exists.

If, as is suggested, this irregularity is similar to that in Coun-
try Gentleman sweet corn, it was probably another set of con-
ditions that caused the trouble. As the writer has since pointed
out,’ the irregularity in the rows of this variety is the more or .Ț
less complete expression of a very definite and comparatively
simple state of affairs. Each female spikelet of ordinary maize

1‘*Tnheritance in Maize,’’ Bull. Conn. Agr. Expt. Sta., 167, 1911, p. 132.

2‘‘ The Morphology of the Flowers of Zea Mays,’’ Bull. Torrey Club, 43
127-144, 1916.

270 THE AMERICAN NATURALIST [Vor. LIII

produces one grain; but in Country Gentleman sweet corn, a
second flower, ordinarily aborted, becomes functional, and the
spikelet produces two grains. Since there is little or no com-
pensation for this in the length of the cob, and insufficient differ-
ence in the size and shape of the grain, the ear is producing a
larger volume of embryo and endosperm than is ordinarily pro-
duced in the same space. As a result of this crowded condition,
the straight rows are more or less obliterated for a more econom-
ical arrangement. At times, however, a set of conditions, pre-
sumably environmental, may limit the size of the grain or
increase the length of the cob sufficiently that the rows are almost
straight, although each spikelet is still producing two grains.
The genetic experiment, then, was probably dealing with an in-
definite expression of a definite characteristic. If the heredity of
the two-flowered condition of the spikelet had been tested, a more
direct explanation would probably have been afforded.

Again (p. 134), these same authorities explain the occurrence
of hermaphrodite flowers upon the basis that ‘‘the immature sex
organs, so-called, of maize seem endowed with the power of be-
coming either stamens or carpels.’’ In so far as actual genetic
results are concerned, this is, in most cases, at least, a sound
working basis, but it is far from exact morphologically. There
is no organ in the young maize flower that has the possibility of
becoming in some cases a stamen and in others a pistil. The
young flower has the ability to become either staminate or pistil-
late because it contains primordia of both stamens and a pistil,
one or the other of which usually does not develop to maturity.®

Blaringhem’s extensive experiments,* in which he attempts to
initiate mutation by means of injuries to the plant, fail to take
into full account certain very significant facts of morphology.
It is probably for this reason that he believes that the acquisition
of hermaphrodite flowers in the maize plant is a progressive step.
On the contrary, every indication points to the fact that the
rudimentary stamens and pistils that have been found in the
flowers of maize are the vestiges of organs that have been, and
not the phylogenetic forerunners of organs that are to be.
Moreover, normal behavior shows that in mutilating the plants
he had merely promoted the production of suckers, which nor-
mally tend to have bisexual inflorescences. Blaringhem’s method
is ingenious and would, no doubt, give good results in a study
of physiology of moncecism; but, the normal plant being under-

3 Ibid., pp. 129-134.

4Blaringhem, L., ‘‘ Mutation et traumatismes,’’ Paris, 1908.

No. 626] SHORTER ARTICLES AND DISCUSSION 271

stood, and full allowance being made for the recognized effects
of inbreeding, it is not believed that there is any clear evidence
that he produced a single new hereditary characteristic in maize.

But not all of the assumptions of fundamentals upon which
geneticists have based their work on maize have been so unhap-
pily chosen as those cited. Most of the work that has been done
on the heredity of endosperm characters depends upon the so-
called ‘‘double fecundation’’ and upon the degeneracy of three
of the four potential megaspores. The former of these facts was
observed by Guignard® in 1901, but he did not figure it; the
latter has been deduced by analogy. Circumstantial evidence
was good in both eases, but evidence of this kind is not always
dependable. No one would risk much in a financial way on
chances like these, but some geneticists have risked years of
work. In a recent paper® the writer has verified the facts as-
sumed in this work.

The peculiar behavior of reciprocal crosses between varieties
of corn differing in the physical nature of the starchy endosperm,
has been explained’? by the assumption that the two hereditary
factors presumably carried by the two polar nuclei be dominant
to the one factor carried by the sperm entering into the consti-
tution of the primary endosperm nucleus. This idea is in accord
with the multiple factor hypothesis, and the phenomenon is one
of the few direct evidences that we have as to the behavior of a
double application of a factor as opposed to a single application
of its allelomorph. But so little is known of the morphology
and the chemistry of these two kinds of starch and their relation
to the surrounding tissues that it is not at all improbable that
the explanation advanced may be modified by the results of
further investigation.

An interesting light is thrown upon the the multiple factor
theory by certain other morphological peculiarities of the grain
of corn. The essential idea of the multiple factor hypothesis, in
a simple form, is that a single visible effect may be due to two
or more factors, only one of which is necessary to produce the
same effect, at least in a limited degree. Little is known of the
relative natures of the two or more factors that compose the mul-
tiple unit in the cases that have been investigated; they may be

5 Guignard, L., ‘‘La double fécondation dans le Mais,’’ Jour. de Bot.,
15: 37-50, 1901.

®**Gametogenesis and Fecundation as the Basis of Xenia and Heredity
in the Endosperm of Zea Mays,’’ Bull. Torrey Club, 46: 73-90, 1919. '

1 Hayes, H. K., and East, E. M., ‘‘ Further Experiments on Inheritance
in Maize,” Bull. Conn. igt Expt. Sta., 188, pp. 12-13, 1915.

272 THE AMERICAN NATURALIST [Von. LIII

alike, or they may be very different from each other. A grain
of corn homozygous for yellow starch and red aleurone is dif-
ferent in color from one having only one of these characteristics.

But to a person with defective vision, or when viewed in a light
of proper color, these two colors and a combination of the two
may appear to be merely different shades of one color. By
breeding this stock with a homozygous white, carrying no con-
flicting factors, we should get what would be to this same defect-
ive vision a perfect illustration of the behavior of multiple fac-
tors. But it is in reality a case of dihybridism in which we have
failed to distinguish between the two sets of allelomorphs. And
who can doubt that relatively as great a lack of discrimination
may characterize our chemical, physical, or morphological vision
in observing some of the classical illustrations to which the mul-
tiple factor hypothesis is applicable?

Other examples could be selected from the work that has been
done on maize, and doubtless many are available from the inves-
tigations made with other plants and with animals, but these will
suffice for illustration. Many of the organisms most useful for
establishing and testing principles of heredity have an external
appearance that may be very deceptive as an indicator of their
true structure, and the true structure alone is the key to the
deeper significance of their genetic behavior.

AUL WEATHERWAX
INDIANA UNIVERSITY

ON HETEROPHYLLY IN WATER PLANTS

THE occurrence of two or more different types of leaf upon one
individual, which is so frequently characteristic of water plants,
has long attracted the interest of botanists. The most usual case
is that in which the submerged leaves are finely divided while the
floating or aerial leaves are relatively simple. Lyte’s Herbal
(1578) contains a vivid description of this type of heterophylly
in the water buttercup. Since this description is also noteworthy
for its insistence on the influence of external conditions, it may
be cited here.

Amongst the fleeting [floating] herbes, there is also a certayne herbe
whiche some call water Lyverworte, at the rootes whereof hang very
many hearie strings like rootes, the which doth Ping gui change his
uppermost leaves according to the places where as it grow That
whiche groweth within the water, carrieth, upon slender milki his
leaves very small cut, much like the leaves of the common — mill,

No. 626] SHORTER ARTICLES AND DISCUSSION 273

but before they be under the water, and growing above about the toppe
of the stalkes, it beareth small rounde leaves, somewhat dented, or un-
evenly cut about. That kind whiche groweth out of the water in the
borders of diches, hath none other but the small jagged leaves. That
whiche groweth adjoyning to the water, and is sometimes drenched or
overwhelmed with water, hath also at the top of the stalkes, small
rounde leaves, but much more dented than the round leaves of that
whiche groweth alwayes in the water.

Among certain Nymphezacee we find a different type of hetero-
phylly in which the submerged leaves are large, thin and translu-
cent, somewhat resembling the seaweed Ulva. These leaves are
particularly well shown in the yellow water-lily.

To enumerate all the varieties of submerged leaf met with
among angiosperms would be too long a task to undertake in the
present paper. It must suffice to say that they arë either highly
divided, ribbon-like, or else thinner and broader than the cor-
responding air leaves. They are characterized anatomically by
the lack of stomates and by the presence of chlorophyll in the
epidermis. They are thus well suited for the absorption of car-
bon dioxide in the dissolved form in which it presents itself to
water plants.

In considering the significance of heterophylly, it is a matter of
importance to remember that the occurrence of different leaf-
forms in a single individual is not confined to aquaties but occurs
also in terrestrial plants. Nehemiah Grew, as long ago as 1682,
pointed out that in many eases one plant bears leaves
of Two Kinds or Two distinct Figures; as the Bitter-sweet, the common
Little Bell, Valerian, Lady-Smocks, and others. For the Under leaves
of Bitter-Sweet, are Entire; the Upper, with two Lobes: the Under
Leaves of the Little Bell, like those of Pancy; the Upper, like those of
Carnation, or of Sweet William.

We find parallels to the heterophylly of hydrophytes, not only
among terrestrial flowering plants, but also in the case of the dis-
tinct ‘‘youth forms”? of conifers, and even—more remotely— in
the ‘‘Chantransia’’ stage of such alge as Batrachospermum.
Heterophylly is indeed so widespread that no interpretation can

valid unless the condition be treated broadly as a very general
attribute of plant life, rather than as a rare and exceptional phe-
nomenon, for which special and individual explanations will
suffice.

To the earlier writers, such as Lamarck, the problem of hetero-
phylly presented no difficulties. They regarded the submerged
or aerial type of leaf as representing a direct response, on the

274 THE AMERICAN NATURALIST [Vor LMI

part of the plant, to the medium. The work of the last thirty
years has, however, rendered this simple conception untenable ;
the theory that now holds the field accords a much less prominent
place to adaptation. The first observation that shook the founda-
tions of the idea that leaf form necessarily depended directly on
the milieu, was that of Costantin, who showed that, in the case of
Sagittaria, the aquatic and aerial leaves were already differen-
tiated from one another in the submerged bud; he noticed: au-
ricles on a leaf which was only 2 to 3 mm. long. In Ranunculus
heterophyllus, also, the leaves destined to be aerial are differen-
tiated in the

A large amount of experimental work has been published by
various authors on the effect of conditions upon the leaf forms of
heterophyllous plants, and, although some of the results are con-
fused and conflicting, a study of the literature seems to justify
one general conclusion—namely, that, in many cases, the sub-
merged type of leaf is, in reality, the juvenile form, but can be
produced later in the life history in consequence of poor condi-
tions of nutrition ; the air leaf, on the other hand, is the product
of the plant in full vigor and maturity. This conclusion, which
is primarily due to Goebel and his pupils, is substantiated not
only by experiments but by observations in the field.

In many heterophyllous plants, the first leaves produced by
a seedling, whether it develops on land or in water, conform,
more or less, to the submerged type. This is the case for in-
stance, in the Alismaceæ. In Alisma plantago, the water plantain,
and Sagittaria sagittifolia, the arrowhead, the first leaves pro-
duced by the seedling (or the germinating tuber) are ribbon-
like, even when the young plant is terrestrial. The formation of
this type of leaf can be induced again, even in maturity, by con-
ditions which cause a general weakening of the plant. Costantin
thirty years ago, recorded that, when the leaves of Alisma plan-
tago were cut off in the process of clearing out a water course, or
in a laboratory experiment, the next leaves produced were rib-
bon-like, thus representing a regression to the submerged form.
More recently, another worker, Wachter, tried the experiment of
cutting off the roots of healthy, terrestrial plants of Sagittaria
natans bearing leaves with differentiated laminæ. It was neces-
sary to cut the roots away every week, as they grew again so
rapidly. The result of this treatment was that the plants were
found to revert to the juvenile stage, the new leaves being band
shaped. When the experimenter ceased to interfere with the
roots, the plants again formed leaves with lamine. Other plants,

No. 626] SHORTER ARTICLES AND DISCUSSION 275

with uninjured roots, grown as water-cultures in distilled water,
also produced the juvenile leaf form, while those grown in a com-
‘plete culture solution developed their laminæ normally.

The same observer recorded a case in which a plant of Hydro-
cleis nymphoides Buchenau (Butomacex), which had been bear-
ing the mature form of leaf, was observed to revert to the ribbon
form. On examination it was found that most of the roots had
died off. When a fresh crop of roots was produced, the mature
type of leaf oceurred again.

Another writer, Montesantos, showed by a series of experi-
ments upon Limnobium Boscii (Hydrocharitacee) that, in this
case also, the heterophylly is not a direct adaptation to land or
water life, but that the floating leaves are ‘‘ Hemmungsbil-
dungen’’ due to poor nutrition. In the water soldier, Stratiotes
aloides, also, he showed that the stomatteless leaves were primary,
but that their production could be induced at later stages by un-
favorable conditions.

An experiment tried by Goebel on Sagittaria sagittifolia indi-
cated that absence of light in this case inhibits the formation of
leaves of the aerial type. An observation of Gliick’s on Alisma
graminifolium Ehrh., also points to the same conclusion. But it
seems probable that the effect produced in these cases was not
due directly to the darkness, but to the state of inadequate nu-
trition brought about by the lack of light for carbon assimilation.

Among the potamogetons, again, experimental work by Esen-
beck has shown that reversion to juvenile leaves can be obtained
under conditions of poor nutrition. For example, when a land
plant of P. fluitans, which had been transferred to deep distilled
water, had its adventitious roots repeatedly amputated, regres-
sion was obtained to the floating type of leaf and then the sub-
merged type. A similar reversion to thin narrow leaves was
brought about in the case of P. natans by growing the upper
internodes of a shoot as a cutting.

Water lily leaves respond to experimental treatment in just the
same way as the monocotyledons already mentioned. In the
case of two species of Castalia, it has been found possible to in-
duce the mature plants to form submerged leaves, either by re-
moving the floating leaves or by cutting off the roots. This con-
firms an earlier suggestion, made by an Italian writer, Arcangeli,
that the development of the submerged leaves of Nymphaea lutea
was due to ‘‘un indebolimento o diminuzione di energia vitale.”
This suggestion has received independent, experimental confir-

276 THE AMERICAN NATURALIST (Vorn: LIIT

mation from Brand, who estimated: that a well-developed floating
leaf of Nymphæa lutea was about eleven times the dry weight of
a submerged leaf of the same area.

Another dicotyledon, Proserpinaca palustris, which was in-
vestigated by Burns, gave experimental results pointing to the
same general conclusion as those already quoted. The primitive
type of leaf in this plant is always a ‘‘water’’ leaf, but this type
of leaf was also produced in the autumn by all the plants, re-
gardless of any external conditions which the experimenter could
eontrol. On the other hand, at the time of flowering and in the
summer generally, almost every plant, whether growing in water
or air, produced the ‘‘land’’ type of leaf—the transition from the
‘‘water’’ to the ‘‘land’’ type taking place earlier on strongly
growing than on weak stems. The author considers it evident
that the aquatic environment is not the cause of the division of
the leaf, nor does it depend on light, temperature, gaseous con-
tent of the water or contact stimulus. The only conclusion which
he holds to be justified by his experiments is that Proserpinaca
palustris has two forms, an adult form and a juvenile form;
under good vegetative conditions, it tends to produce the adult
form with the undivided leaf, the blossom and the fruit, while, if
the vegetative conditions are unfavorably influenced, a reversion
ean be induced to the primitive form with the submerged type
of leaf. These results are consistent with those of McCallum,
who had dealt with the same species at an earlier date, but his in-
terpretation is slightly different. He is inclined to regard the
occurrence of the water form as induced by the checking of trans-
piration and the increased amount of water which hence accumu-
lates in the protoplasm. This explanation is not inconsistent
with the more general view that any condition tending to lower
the vitality may be responsible for a reversion to the submerged
type of leaf.

In nature, the regression to the juvenile type of leaf sometimes
occurs, not only in the case of an entire plant subjected to adverse
conditions, but also in the ease of lateral shoots from an individ-
ual which is otherwise producing the mature form of leaf.
Goebel for instance, examined an old example of Eichhornia
azurea (Pontederiacee) which had wintered as a terrestrial plant
in a greenhouse; the leaves were of the mature form, differen-
tiated into sheathing base, petiole and lamina, except in the case
of a lateral shoot, which, on the contrary, bore the grass like,
simple leaves which characterize the young plant. Goebel also

No. 626] SHORTER ARTICLES AND DISCUSSION 277

describes the occurrence of subdivided leaves of the water type
on lateral shoots of normal land plants of Limnophila hetero-
phylla. A corresponding reversion has been observed in the case
of the side branches of plants of Proserpinaca palustris develop-
ing in the air from a plant whose main stem was producing the
mature type of leaf; by removing the growing apex of the stem
in June, side branches of the ‘‘water type’’ were induced to de-
velop.

The interest of these lateral shoots, which show a reversion to
an ontogenetically earlier type of leaf, is enhanced by the fact
that C. and F. Darwin in ‘‘The Power of Movement in Plants”?
have recorded a case of the occurrence, on lateral shoots, of leaves
whose characters are probably phylogenetically earlier than those
which the species normally exhibits. Their observations related
to the sleep habits of the allied genera, Melilotus and Trifolium.
They noticed in Melilotus Taurica that leaves arising from young
shoots, produced on plants which had been cut down and kept in
pots during the winter in a greenhouse, slept like those of Tri-
folium, with the central leaflet simply bent upwards, while the
leaves on the fully grown branches of the same plant afterwards
slept according to the normal Melilotus method, in which the ter-
minal leafiet rotates at night so as to present one lateral edge to
the zenith. They suggest that Melilotus may be descended from a
form which slept like Trifolium.

The idea that the ‘‘juvenile’’ leaves produced on lateral shoots
may in some cases represent an ancestral type, is consistent with
the facts in the case, for instance, of the Alismaceæ, provided that
the ‘‘phyllode theory’’ of the monocotyledonous leaf be accepted
in the sense advocated by Professor Henslow and the present
writer. According to this theory, the ancestral leaf of this family
was ribbon-shaped, while the oval or sagittate blade (or ‘‘ pseudo-
lamina’’) represents a later development—a mere expansion of
the apex of the petiole. The submerged youth leaves of this fam-
ily would thus represent a reversion to phylogenetically older

orms.

If the interpretation of heterophylly indicated in the present
paper holds good at all widely, the teleological view of the sub-
merged leaf must be considerably modified. The present writer
would like to suggest that, for the old conception of heterophylly
as induced by aquatic life, we should substitute the idea that such
a difference between the juvenile and mature forms of leaf as
would render the juvenile leaf well suited to aquatic life, has been

278 THE AMERICAN NATURALIST [Von. LITI

in many cases one of the necessary preliminaries to the migration
from land to water, and that the aquatic angiosperms thus in-
clude, by a process of sifting, those plants whose terrestrial an-
cestors were endowed with a strong tendeney towards hetero-
phylly.
AGNES ARBER
NEWNHAM COLLEGE,
CAMBRIDGE

COALESCENCE OF THE SHELL-PLATES IN CHITON*

CHITONS are peculiar in the fact that the molluscan shell is
here represented by a series of eight distinct dorsal plates, which
in different genera overlap and articulate with one another to
varying degrees. The full number, 8, seems, however, to be
invariably present. While examining recently a series of some-
what over 2,100 individuals of Chiton tuberculatus L., I came
upon two cases, and two only, exhibiting any irregularity with
respect to the number of the shell-plates. These were specimens,
a male and a female, found near together on the beach at Cross
Bay, Bermuda, in which plates 7 and 8 had in each instance
almost completely fused (Figs. 2-6), so that each of these ani-
mals seemed at first sight to have but 7 plates; since no rec-
ords seem previously to have been made of such occurrences,
they are here figured and described.

In the two abnormal chitons the fused terminal plates were
of similar external appearance, but in individual A, the female,
the coalescence of plates 7 and 8 was somewhat less complete
than in individual B, the male, as shown by the form of the
inner surfaces of the compound plates. It is perhaps accidental
that in both cases fusion of the respective plates is somewhat
assymetrical, being more complete on the right side. As seen
in Fig. 4, the muscular intersegmentum, which ordinarily re-
ceives the insertion plates of the eighth valve, is represented by
a relatively small tongue of tissue.

1 We owe to Dr. H. B. Guppy, F.R.S., the important idea that the habitats
of plants are determined by their ‘posuere ies of structure and not
vice versa. In relation to the occurrence of plants with buoyant seeds
and fruits in water-side stations, he writes, ‘‘there are gathered at the mar-
gins of rivers and ponds, as well as at the sea-border, most of the British
plants that could be assisted in the distribution of their seeds by the agency
of water. This great sifting experiment has been the work of the ages, an
we here get a glimpse at Nature in the act of selecting a station

* Contributions from the Bermuda Biological Station for Research, No. 104.

No. 626] SHORTER ARTICLES AND DISCUSSION 279

It may be of significance that the only instances obtained of
fusions of the kind figured, occurred at a sandy beach, on the
south side of Bermuda, exposed to the beating of the ocean
surf. Individual A, when found, was attached to a rock, but
was half-covered by sand left by the tide. Chitons in such sit-
uations are frequently buried for a time beneath a foot or more
of sand, and under these circumstances the over-lapping edges

Outlines of valves 7 and 8 of a normal Chiton tuberculatus; a, in

Ae
dae hamra relations ; b, plate 8 separately. Natural size.
Compound terminal plate of an abnormal C. tuberculatus (individual

4. Compound terminal plate of an ashnormeal 0. tuberculatus (individual
B, 2 E cm. long; dorsal view. Natural size
The same, ventral view. Natural stie.
we ‘ Dorsal aspect of g aetan] end of fgg A, to show (8) reduction
of intersegmentum 7-8; (a) intersegmentum 6-7. Both abnormal chitons esti-
mated to be five years old. Natari size.

of the shell-plates are kept tightly pressed together, thus pre-
venting sand-grains from abraiding the soft inter-tegmental
mantle. The posterior end of a Chiton tuberculatus is less
active in turning movements, in curling-up and in similar opera-
tions than is the anterior end, so that two valves, once stuck
together, might, at the posterior end, have a better chance of
remaining together. The incomplete union of the valves, visible
when seen from their inner side, suggests that the coalesced
plates started out independently. Whether or not this view be
valid, it would be of interest to determine if there is any gen-
eral tendency, in special localities, toward the establishment of
races of chiton possessing a reduced number of plates.
W. J. CROZIER

DYER ISLAND, .
BERMUDA

280 THE AMERICAN NATURALIST [Von. LIII

THE EFFECTS OF THE WINTER OF 1917-1918 ON THE
OCCURRENCE OF SAGARTIA LUCIÆ VERRILL!

In June, 1902, I published in the AMERICAN NATURALIST some
notes on the dispersal of Sagartia lucie that tended to show
that this sea-anemone had spread from the neighborhood of
New Haven, Conn., along the New England coast as far north
as Salem, Mass. This migration was accomplished in approxi-
mately a decade, from 1892 to 1901. Since 1902 repeated ef-
forts have been made to discover evidences of this species farther
to the north than Salem but without avail. Apparently the
species had reached its northernmost limits.

Sagartia lucie was first noticed in Woods Hole, Mass., in
1898. From that year until the present it has been an extremely
abundant species on the stones, mussels and eel grass in the
waters of this region. On Pine Island, a narrow ridge of rocky
gravel overtopped with coarse vegetation and lying in the swift
tidal currents of the Hole, the narrow beaches between tides
have been covered with thousands of this species of sea-anemone.
When this locality was visited in June, 1918, not a single speci-
men of Sagartia lucie could be found, though the particular
area examined had been covered with many individuals the
year before. Nor was this condition due to the relatively early
date at which the search was made. Repeated attempts during
low tides in July and August never yielded at Pine Island
more than two or three specimens at a time, and it was quite
clear that Sagartia lucie, once so prevalent in that locality, had
suddenly become all but extinct there. The same was true of
other situations in and about Woods Hole. In fact, a general
search showed that in not a single location where this sea-
anemone had been abundant in 1917 could there be found more
than a paltry number of specimens in 1918

The occasion of this sudden and great diminution in the num-
bers of Sagartia lucie is to be attributed, I believe, to the rigor
of the winter of 1917-1918. The cold and ice of this winter
were almost unprecedented. Mr. Vinal Edwards, the veteran
collector of the laboratory of the United States Bureau of Fish-
eries at Woods Hole, has kept a continuous record of the weather
conditions of this region for a long period and this record
shows, as might be expected, that the winter conditions in 1917-
1918 were more severe than for many years past. In no win-

1 Contributions from the Zoological Laboratory of the Museum of Com-
parative Zoology at Harvard College.

No. 626] SHORTER ARTICLES AND DISCUSSION 281

ter during the last ten years has the sea water been at 0° C. or
lower for so long a period as last winter. Beginning with the
season of 1908-1909 and proceeding to that of 1917-1918, the
number of days for each of the ten winters in which the tempera-
ture of the seawater was O° C., or lower, was 3, 40, 44, 63, 3, 55, 0,
65, 36 and 80. Thus 1917-4918 with its 80 days of extremely
cold water strikingly outruns any one of the preceding nine
years.

This winter was conspicuous for the formation of large
amounts of anchor frost in the shallow waters about Woods
Hole. This frost or ice can be seen forming on the bottom of
paion bodies of salt water when the temperature of that water

at 0° C., or lower. It is apparently due to the freezing of
real cater that, seeping through the land, rises from the sea
bottom and solidifies at once on coming in contact with sea-
water below its own freezing point. This fresh-water ice is
especially destructive to marine animals on the bottom and its
great prevalence during the winter of 1917-1918 is probably re-
sponsible for the scarcity of sea-urchins and other like forms the
following summer. It probably had little or no effect on Sagartia,
for this sea-anemone lives chiefly between tides and, therefore,
above the level at which anchor frost is found, but as a winter
phenomenon this ice is a good index of severity and it is severity
in the nature of low temperature that is responsible, I believe, for
the almost complete elimination of Sagartia.

That this sea-anemone was not destroyed by the merely me-
chanical effect of ice and waves is seen from the fact that the
same stretches on Pine Island that were populated with Sagartia
lucie were, and still are, covered with many specimens of
Metridium marginatum. This northern species seems not to
have suffered in the least from the severity of the past winter
and I, therefore, conclude, since Metridium was as much ex-
posed to mechanical injury as Sagartia and still survived m
ordinary numbers, that Sagartia succumbed to low temperature
rather than to any other factor in its environment. This is in
accord with the general belief, originally expressed by Verrill,
that Sagartia lucie is a southern species introduced by some
accident into northern waters. Granting this conclusion, lit is
easy to understand why this species has not migrated farther
northward into colder waters and why in severe winters it is
almost exterminated in localities such as Woods Hole.

G. H. PARKER
HARVARD UNIVERSITY

282 THE AMERICAN NATURALIST [Vou LIII

TAXONOMY AND EVOLUTION
A REJOINDER

THE writer has great sympathy with much of what ‘‘X’’ has
to say on the above subject in a recent number of the AMERICAN
NaTuRALIST (Vol. xuvm, 369-382).! Needless to say, however,
he can not agree with all. True there is much in systematic
zoology that is slipshod, but till statistics can be produced to
show that the percentage of slipshod work produced by sys-
tematice zoologists is higher than in other fields of zoology, the
writer of this article has a temporary residence in Missouri.
He is of the opinion, also, that as great a percentage of the work
of the systematic zoologist will stand the test of time as the work
of the anatomist or any other worker in the field of zoology and
proposes to remain of that opinion until time, the great leveler,
proves to the contrary.

Linnzus is apparently not the only genius that has left the
back door open and that has ‘‘been followed by a crowd of other
workers eager to attain to immortality,” as witness the great
mass of half-digested literature on genetics, say, that has been
crowded into the past ten or a.dozen years. It would be a sad
state of affairs indeed if systematics as a whole were not im-
proving. That there have been occasional backward steps there
is no doubt, but on the whole the progress has been forward.
I hardly believe that even the systematists are as big fools as
‘*X”’ pictures them to be, for I have yet to discover in my
rambles a systematist who believed that his work was final.
Heaven forbid. The ezar in zoological nomenclature may arise
and issue his fiat, but there will be later czars who will do away
with them. For surely ‘‘X’’ would not have us believe that the
day will ever dawn in this world when all things are settled.
My shorter catechism is somewhat awry, but surely such a happy
state is reserved for the Great Beyond.

Without wishing to disparage the modern workers T wish to
say that some of the older workers did write ‘‘careful descrip-
tions,’’ as witness the following case which has been called to my

1The present paper was written in July, 1914, soon after reading the

r by ‘*X.’’ It was laid away but now that it is more than four years
old ‘‘ going on’’ five, as children say, it seems best to submit it for publi-
eation.

No. 625] SHORTER ARTICLES AND DISCUSSION 283

attention. One of the early systematic entomologists described
a species, on external characters only, in about three lines.
Later entomologists were puzzled because the species had charac-
ters common to two widely separated genera; and one systematist
said it belonged to one genus, and another said to a widely
divergent genus, while a third said it was simply another name
for a common form. Yet, behold, when the species was redis-
covered it was found to belong to a new genus with characters
common to the two widely divergent genera. Now, what’s the
answer, certainly the original description must have been a good
one otherwise how could workers nearly a century later rec-
ognize the characters?

Isolated quotations from descriptions of any species look
ridiculous (p. 370), but no more so than isolated quotations
from the work of sedli neurologist or what not. A kindly
feeling for my fellow workers in other fields and for the editor
of the AMERICAN NATURALIST stays me from quoting at length
and verbatim. Fortunately ‘‘X’’ has sufficiently concealed his
identity so that I can not quote some of his own discussions
until he yawns. Neither is my soul more deeply stirred by con-
templating the poor hymenopterist, squinting at his box of dried

‘‘bugs’’ stuck on pins; than it is by the poor hunch-backed short-
sighted cytologist (let us say) who, peering through his high
power compound microscope, imagines that the world is circum-
scribed by his field of view and that a cell, or a nucleus, or a
chromosome, is all there is to zoology.

**X’’ seems to deplore the fact of specialization in zoology and
at the same time seems to ignore the fact that it is along these
lines that the world moves. Why should we not have neurolo-
gists, taxonomists, hemipterists, ete., in zoology just as we have
masons, carpenters, roofers, pinata, tinners, ete. How many
railroads would have been built in this world or how much
progress would have been made in any other line of human en-
deavor if every man had to be a jack-of-all-trades? Do we hire
a man to build us a house? Most certainly not. We hire a
brick mason to lay the foundation, a carpenter to erect the
frame, another one to put on the weather boarding, and still
another to do the finishing inside; and so on until our house is
finished and the whole structure stands only as long as the work
of each one of these individual workers will stand. So it seems
to me it is in zoology, the systematist lays the foundation upon

284 THE AMERICAN NATURALIST [Vor. LIII

which the whole structure is raised. And while the whole
method of systematice zoology is open to criticism by anatomists,
or what not, yet a certain amount of systematic work must be
done before the anatomist can develop his work. If we take
this position it seems to me that we must grant that the sys-
tematist must be far to the forefront, well in advance of the
workers in other fields. And certainly this much must be said
in his favor that he has turned out enough ‘‘new species’’ in
the last few years to keep the rest of the zoologists busy for a
year or two.

‘*X’s’’ whole attitude is that the systematist makes mistakes
and that he sticks only to external characters. In regard to the
first I would call ‘‘X’s’’ attention to the fact that anatomists a
little less than 300 years ago believed the arteries carried air, not
blood. And it seems to me if we go back about 250 years we find
one Robert Hooke describing ‘‘little boxes (empty) of cells dis-
tinct one from another’’; and wasn’t it only about half a century
ago that the cytologist awakened to the fact that the boxes were
not as empty as might seem? Now the question to my mind is
this, would we know as much about cytology as we know to-day,
if Hooke had not discovered his empty boxes? I think not.
And as a necessary corollary would we know as much about the
animal world as we now know if systematists had not described
new species? I think not. The fundamental basis of systematic
work, it seems to me, must always be external characters, though
they may be variable and unsatisfactory in many respects.
What we all want and what I believe all systematists are striving
for though some of their strivings may be misdirected is, among
other things, ease of identification which, to my mind, implies
reference to external characters. I, see a woodpecker sitting in
a tree and identify him as a yellow-bellied sapsucker by the fact
that he has, among other characters, a white stripe down his
wings. Very unscientific, I grant, but highly satisfactory to me
if I am collecting not sapsuckers but downy woodpeckers. Also
to the sapsucker if the alternate character which enabled me to
identify him was the presence of extra small convolutions on his
cerebellum,

I make my plea for systematic zoology as systematic zoology,
not for its ‘‘phylogenetie classification of animals,’’ nor for its

on geographical distribution, variation or heredity or
sarsii else. The description of ‘‘560 new species of Zonitide’”’

No. 625] SHORTER ARTICLES AND DISCUSSION 285

may not seem soul-inspiring work to ‘‘X,’’ but to the deseriber
it may have been exceedingly so. The description of 560 new
species of Zonitide makes it possible for some student of va-
riation or of ‘‘ phylogenetic classification’’ to work on the Zonitidee
in a way that would not have been possible if these 560 ‘‘new
species’’ had not been described, and no one man would have
been able to describe the 560 new species and work their em-
bryology, internal anatomy, neurology, ecology, geographical dis-
tribulation, behavior, variation, mendelian relations, etc., and live
to tell the tale.

Furthermore, if there is any man that has the aptitude to
describe ‘‘560 new species of Zonitidse’’ my benediction is ‘‘let
him go to it.’’ And while 585 of his ‘‘560 new species’? may
prove to be false alarms that have never been turned in, at the
same time it does not seem to have occurred to ‘‘X’’ that he may
be doing much less harm thusly employed than if, he were
rampant with scissors, scalpels and needles or with killing
agents, stains and a microtome trying to discover the true in-
wardness of the Zonitide. I do not want to be misjudged by
any one who may think that I am making a plea for slipshod
work, but I do want to make a plea for the isolated worker who
is plodding away in his own particular field without hope of
reward or recompense in this day or generation. Let us be very
careful about setting our stamp upon a thing as worth while
or not worth while. Mendel, the poor isolated monk, working
away with his peas, never dared dream, I venture to say, that
his work would revolutionize the biological thought of the
twentieth century. Thus ‘‘X’’ may have the misfortune to view
in a future reincarnation the sad spectacle of the zoologists of
say 200 years hence loudly acclaiming the good work of the
describer of the 560 new species of Zonitide, while at the same
time they point with scorn to the work of the anatomist who
discovered (?) that the digestive system of the Zonitide runs
up hill.

The writer has the fortune or misfortune, as pleases your
point of view, to be the entomologist of a state experiment
station. His principal duties as entomologist are the intensive
studies of two widely separated species of extremely injurious
insects. This work is carried on under the Adams Fund by
grants from the United States Department of Agriculture. Both
projects were so outlined as to involve everything about these

286 THE AMERICAN NATURALIST [Vou. LIII

two insects that could be discovered by the writer; internal and
external anatomy, embryology, life history, parasites, ete.
Present indications are that it will take an average of about
six years to finish (?) each one of these projects. Yet such a
seemingly slow rate of progress is made possible only by the
fact that some one working somewhere has described these two
species and given them names. The one species was described
without the describer ever having seen the male! Yet without
this inadequate description progress on this problem would
have been very greatly delayed. And so it is in every other
field that these problems touch. Some one has described some-
where 29 species of parasitic hymenoptera, one of this number
preying upon one of the species involved. Yet the describer
knew only the adult and that only imperfectly, but his knowl-
edge plus my own sends us one step nearer the complete knowl-
edge of this species which ‘‘X’’ craves. And our knowledge of
this species plus some one’s knowledge of other related species
raises us just one step nearer the truth which should be the
goal of all human thought, and all science, zoology not even
excepted.
_ I am interested in the phylogenetic relationship of a group
of insects of no great economic importance. Especially am I
interested in the genealogical tree of these insects as shown by
the groups of characters of one structure. Now such work is
made possible because three men in this country have devoted
their entire time describing new species and new genera in this
group. Without these descriptions many of which might have
served as well as the one quoted on page 370, and without the
collections of insects which these three men have made it would
be impossible for me to make any progress along the line of a
genealogical tree, which it is my fond hope will be of some use
to the systematists of this group and to zoologists in general,

I have long wondered what could be called trivial characters.
A few illustrations of the importance of so-called trivial char-
acters in other fields than systematics may perhaps occur to ‘‘ X.”
One of the most important that has come to my notice was that
of a cytologist who discovered differences in the chromosomal
characters of two different sets of individuals of the ‘‘same
species’’ only to discover later that systematists had long dis-
tinguished between these two forms on the basis of characters
more trivial than whether they were ‘‘pink with blue spots” or

No. 625] SHORTER ARTICLES AND DISCUSSION 287

‘“blue with pink spots.’’ Again two species of scale insects
are separated by the fact that one has the median lobes of the
pygidium rounded while the other has the median lobes conical
(external characters). Yet one lives on oak trees and has at
least four generations annually and the other lives on maple trees
and has only a single generation annually. Now if ‘‘X’’ thinks
that these facts would have been discovered as easily and as
quickly as they have been discovered, if Professor Comstock had
not pointed out these ‘‘trivial characters’’ some thirty-odd years
ago, he thinks differently than I think. Yet the application of
these facts is of vast importance to the horticulturalist and land-
scape architect or any other artisan who works to beautify our
landscapes with trees, or any one who attempts to control these
two pests.

I have no doubt that Linneus was accused of relying on
trivial characters for separating some of his genera and species.
It would be interesting if history could tell us and it would be
still more interesting if we could look into the future, say 100
years, and see what systematists and others will say about the
present-day systematists who overlooked such perfectly obvious
characters as the extra spines on the hind leg of species ‘‘a’’
contrasted with ‘‘b’’ and their wonder and amazement that
systematists of this our glorious twentieth century should have
overlooked such important and obvious characters. So it will
be in other fields. The histologist of the future will wonder why
we used such crude killing and fixing agents; and will, more
than likely, refer to our finest precision microtomes with a
shrug much as we refer to the stone hatchets of the men of the
Old Stone Age. |

I make this somewhat extended plea because it seems to me
that ‘‘X’’ has unconsciously done the systematists a great
wrong.* ‘‘X’s’’ attitude may discourage promising young men
from entering the field of systematics where their help is greatly
needed. Let us therefore lay aside our critical air and our
sitting in judgment to decide just what is worth while and what
isn’t and turn our attention to utilizing the results of other
workers in other fields to the greatest extent. The systematist

2 That all may see that my plea is entirely unselfish, I will state that I am
not a systematist and that I have never described a single ‘‘new spe-
cies.’’? My attitude is simply one of gratitude to the systematists who have
helped me with my problems.

288 THE AMERICAN NATURALIST [Von LIH

is human like the rest of us, he has his limitations like the rest
of us, but he believes, I think justly, that his work is pioneer
work of great importance; and, if occasionally he gets beyond
the limited range of our embryologist’s microscope or our an-
atomist’s scalpels and needles, let us not accuse him of wandering
along the River of Doubt or being a lineal descendant of the
famous Baron Munchausen. But let us look upon the sys-
tematist’s work as the foundation for the glorious structure,
modern zoology, which completed by his other co-workers will
stand four square to the wind for all time to come. We do not
need to defend systematics on the basis of ‘‘(1) the advertise-
ment theory; (2) the recognition mark theory ;’’ although both
are perhaps more important than ‘‘X’’ intimates. But what is
vastly more important is the fact that systematics is the basis
for all real work in zoology. And the morphologist or anatomist
who takes the attitude that systematics is to be entirely avoided?
or, what is worse, is to be simply laughed at is placing himself
in the same class as the man who says that there is no such thing
as matter in the world. Sooner or later he is going to bump into
the fact that systematics must play its part in his field and that
systematics is broader than the question whether the ‘‘second
joint is longer than the third’’ or whether a species should be
called aabus Smith or beabus Jones.
Z.

3 Just what does ‘‘X’’ think about the anatomist who discussed at great
length the anatomy of, let us say, the ‘‘ American frog (Rana temporaria) ’’
because that was the name given the frog in his perfectly good English
‘*Text-book of Zoology’’; when the context shows that the frog he was
dealing with was the common leopard frog (Rana pipiens)?

THE
AMERICAN NATURALIST

VoL. LIII. July-August, 1919 No. 627

ON THE USE OF THE SUCKING-FISH FOR
CATCHING FISH AND TURTLES: STUDIES
IN ECHENEIS OR REMORA, II.

E. W. GUDGER

AMERICAN MUSEUM OF NATURAL History, New York Crry

THe FısaerMmaNn-Fısau 1x MozaMmBIQUE WATERS

Ix the year 1884, Mr. Frederick Holmwood, British
Consul at Zanzibar, by publishing an article in the Pro-
ceedings of the Zoological Society of London, brought
this extraordinary use of this remarkable fish to the at-
tention of the scientific world. Chancing on this article,
I became greatly interested in the matter and have been
led to collect all the available data and to present it
herein to those who may be interested.

On a trip in a steam launch from Pemba to Zanzibar,
Holmwood had his attention called to a number of remo-
ras which were attached to the sides and bottom of the
boat. To these the natives on board gave the name
**Chazo.’’ Later at Zanzibar he saw natives digging
out diminutive canoes, too small for any normal use,
which he was told were for the ‘‘Chaza’’ (so he under-
stood the native word). Now ‘‘Chaza’’ is the word for
oysters or other bivalves, hence- he thought that these
were used to gather such in, but his servant told him
that it was a ‘‘house’’ for the ‘‘Chazo’’ or sucking-fish
kept by most fishermen in their huts. Later he learned
that the native fishermen use the Chazo fish to catch
turtles and large fish of any kind. And later still in

289

290 THE AMERICAN NATURALIST [Vou. LIII

Madagascar he was informed that sharks and even large
crocodiles were caught by the use of a fish called
Tarundu' which was trained for the purpose. Unfor-
tunately, just here Holmwood gave vent to his ineredu-
lity and his informants being greatly incensed refused
to talk with him further on this matter.

Holmwood spent considerable time in gaining the con-
fidence of the native fishermen of Zanzibar and was re-
warded by being allowed to visit their huts and examine
the ‘‘Chazo.’’? These he found to be remoras (echeneis?)
from 2 to 4.5 feet long and from 2 to 8 pounds in weight.
They were kept in the little canoes in the cabins and were
so tame as readily to come to the surface of the water at
the appearance of their masters, by whom they allowed
themselves to be freely handled.

Each Chazo had a strong iron ring or loop fixed just above the tail
[text-figure 1] for the purpose of attaching a line to when being em-

TEXT-FiGURD 1. Tail of sucker-fish with band and-ring. (After Holmwood, 1884.)

ployed in hunting. In some cases these appendages had evidently re-
mained on for years, during which the fish had so grown that the iron
had become imbedded in a thick fleshy formation. In two instances the
ring had been inserted in the muscular substance at the root of the tail
[fin], but generally a simple iron band was welded around the thinnest
part of the body a few inches from the tail, which kept it from slipping
off. To this was riveted a small movable ring or loop resembling that
of a watch-handle. In one ease [text-figure 2] this loop was fastened
on by servings of brass wire in a similar manner to the rings of a
fishing rod.

1 Every effort has been made to trace down the use of the Tarundu, but
books on the fishes of Madagascar are few, and none of them nor the works
of travel consulted have given any clue.

No. 627] STUDIES IN ECHENEIS OR REMORA 291

Holmwood purchased one of these fish to send to Eng-
land but it was killed by a crane. A second one died,
probably from lack of a fresh supply of water. He after-
wards arranged to buy another on its return from a fish-
ing trip.

It was brought to me a few weeks later minus its ring, and with a
large wound or rent above the tail, part of which was gone. The owner
declared that it had caught two turtles, which he showed me lying in
his canoe, and that it had afterwards affixed itself to a large shark and,
holding on after all the spare line had been paid out, the tail had given

TEXT-FIGURE 2, Tail of sucker-fish with loop and servings.
(After Holmwood, 1884.)

way. He stated that the Chazo had then relinquished its hold and re-
turned in its mutilated state to the boat. He assured me that this was
not an unusual oceurrence and that after a time a fresh ring would be
attached and the fish become as useful as before. I endeavored to pre-
serve one of these Chazos in spirits of wine, but failed owing to the
inferior quality of the spirit. This specimen measured 2 feet 8 inches
in length and weighed 314 pounds. The sucker contained 23 pairs of
lamelle.

Holmwood wanted to go out with the fishermen and
see the fishes at work. But as the distance to the fishing-
grounds was considerable, as the trips lasted fifteen days,
and as the boats were small and lacked accommodations
for a European, he was forced to desist. Thus he failed
to become an eye witness to this remarkable procedure.?

2 Under date of 1883, a writer signing himself Phil. Robinson published
a pamphlet entitled ‘‘ Fishes of Fancy—Their Place in Myth, Fable, Fairy

Folk-Lore.’? This was issued as a hand book for the great
International Fisheries Exhibit of that year in London. In this is a ver-
batim quotation from an article by Holmwood on the use of the fisherman
fish in the official catalogue of the exhibition. After much difficulty this
official catalogue was located and in it was found Holmwood’s origi

292 THE AMERICAN NATURALIST [Von. LIII

Holmwood’s interesting account is however not the
first for the use of the living fish-hook in Mozambique
waters. In the year 1829 Lacépède published his ‘* His-
toire Naturelle des Poissons,’’ in which, with reference
to foreign fishes, he largely made use of the manuscripts
of the lamented naturalist, Commerson. On page 490 of
Tome III we read:

Commerson ... has written that this fish (Hcheneis naucrates) fre-
quents very often the coast of Mozambique, and that near to this coast
it is employed for fishing for marine turtles in a very remarkable
manner, due to the power which the Echeneis possesses of sticking to
them. We think that we ought to report here the data which Commer-
son has collected on this subject so very curious, the only of the kind
which has ever been observed. [ ?]

There is attached to the tail of the living Naucrates a ring of diameter
sufficiently large not to incommode the fish, and small enough to be re-
tained by the caudal fin. A very long cord is attached to this ring.
When the Echeneis has been thus prepared, it is placed in a vessel full
of salt water, which is renewed very often, and then the fishermen place
this in their boats. They then sail towards those regions frequented by
marine turtles. These animals have the habit of sleeping at the surface
of the water on which they float, and their sleep is so light that the least
noise of an approaching fishing-boat is sufficient to wake them and
cause them to flee to great distances or to plunge to great depths. But
behold the snare which they set from afar for the first turtle which they
perceive asleep. They put into the sea the Naucrates furnished with
its long cord. The animal, delivered in part from its captivity, seeks to
escape by swimming in all directions. There is paid out to it a length
of cord equal to the distance which separates the sea turtle from the
boat of the fishermen. The Naucrates retained-by the line, makes at
first new efforts to get away from the hand which masters it. Soon,
however, perceiving that its efforts are in vain, and that it cannot free
itself, it travels around the circle of which its cord is some fashion a
radius, in order to meet with some point of adhesion and consequently
to find rest. It finds this asylum under the plastron of the floating
turtle, to which it attaches itself easily by means of its buekler, and

account. He wrote up for this an account of the fisheries - Zanzibar and
TERE by giving a short description of fishing with the ‘‘chazo.’’ This

n very abbreviated form the data included above, and ends with the
natant ‘*T hope to forward a specimen of this interesting fish before the
close = exhibition.’’ However, as indicated previously he was unable
to do

No. 627] STUDIES IN ECHENEIS OR REMORA 293

gives thus to the fisherman, to whom it serves as a fulerum, the means of
drawing to them the turtle by pulling in the cord.®

This account of Commerson-Lacépéde’s is very circum-
stantial and exceedingly interesting, but it is not the first
account of the fisherman fish, and not even the first for
East African waters, for in 1809 and 1810 Henry Salt
under orders of the British government made a voyage
to Abyssinia by way of the Cape and the Mozambique
Channel, stopping at Masuril, a village on the harbor of
Mozambique. Of this visit he says under date of Sep-
tember 9, 1809 (his book was published in 1814):

As he [the Bishop of Masuril] was aware of my wish to collect the
rarities of the place, he made me a present . . . of a large sucking-fish
(Echeneis naucrates) . . . which had just been brought in by a fisher-
man. All the Portuguese gentlemen, whom I conversed with on the
subject, agreed in assuring me that fish of this kind were employed on
the coast in catching turtles. The mode of doing this is by confining the
fish with a line to the boat, when it is said invariably to dart forwards,
and to attach itself by its sucker to the lower shell of the first turtle
found in the water, which prevents its sinking, and enables the fisher-
man to secure his prey. The reason for the fish fastening on to the
turtle is supposed to be done (as the Bishop observed) with a view to
self-preservation, and its strength is so great that, when once fastened,
the turtle is rarely known to escape.

Earlier still (in the latter half of the eighteenth cen-
tury) a Swede named Andrew Sparrman made a voyage
to the Cape of Good Hope, and in that part of his book
dealing with the land of Natal, in the French translation
published at Paris in 1787 he wrote:

They [the inhabitants of the country] carry on a very singular
method of fishing for turtles. They take alive a fish called Remora,
and fixing two cords, one to its head and the other to its tail, they then
throw it into the depths of the sea in the region where they judge that
there ought to be turtles, and when they perceive that the animal has
attached itself to a turtle, which it soon does, they draw in to them the
Remora and with it the turtle. It is said that this manner of fishing is
also carried on in Madagascar.

The same account in brief form is found on pages 170-171 of Pasfield
Oliver’s life of Commerson (1909).

294 THE AMERICAN NATURALIST [Vou. LIII

This account is not found in the English. translation
of Sparrman’s voyage, and I have not had access to the
original Swedish edition, but it is found in the French
edition in a sort of appendix to that section describing
South Africa and is credited to Middleton’s ‘‘Geog-
raphy.’’ Inspection of volume I (1777) of. this latter
work revealed the account substantially as given above,
but in quotation marks with no hint whatever of its ulti-
mate source.

Humboldt (1826) refers to a similar incident related
by Captains Dampier and Rogers. Dampier was worked
through twice without finding the reference, but a third
going through his ‘‘ Voyages’’ page by page revealed it
as an annex to part 3 of his volume III, ‘‘A Discourse of
Winds,’’ etc. (6th edition, 1729). Middleton has copied
it almost word for word, so it need not be repeated here.
It will be of interest, however, to note that Dampier says
that this ‘‘annexed paper’’ was ‘‘received from my inge-
nious Friend, Capt. Rogers, who is lately gone to that
place (‘Natal in Africk’): and hath been there several
times before.’”*

It must be remembered that Holmwood wrote of a fish
called Tarundu used in Madagascar as a living fish hook,
and Lacépéde quotes Commerson that a sucking fish is
so used in the Isle of France as well as in the Mozambique
country and lastly that Dampier quotes Rogers as to this
use also in Madagascar. Acting on these hints a good
deal of time has been spent in hunting for such accounts
not only in books on the fishes of these islands but also
in books of travel and at this writing three corroboratory
accounts have been found. The first is to be found in
Pollen’s work on the fisheries of Madagascar (1874).

4 The index to Rogers’ book (‘‘A Cruising Voyage around the World,’’
1726) does not contain the words echeneis, remora or sucking-fish. Careful
examination of the book, and a minute inspection of that part relating to
South Africa, gave no results whatever. Dampier’s ‘‘ Voyages’’ show that
he was keenly observant of natural history objects wherever he went, while
Rogers paid little or no attention to such matters. It seems likely that the |
foregoing account was communicated to Dampier by word of mouth or by
letter from Rogers.

No. 627] STUDIES IN ECHENEIS OR REMORA 295

For Malagassy waters he quotes the use of Hcheneis as
given by Middleton, Commerson-Lacépéde and Salt, an
for other waters other authors to be referred to later.
He is not clear as to its use in his own time but he seems
to indicate that in his day it was so used.

` Our next reference is dated 1897. In the Antananarivo
Annual for that date (published by the London Mis-
sionary Society at the capital of Madagascar) there is
under ‘‘Natural History Notes’’ a translation by James
Wills of a native manuscript which reads as follows:

In the sea off the northwest coast of Madagascar a fish is found
called by the people Hamby. It is round and long, somewhat like a
lizard, but its tail unfolds for swimming like that of a gold-fish, and it
has fins on each side. The length of a full-sized one is about that of a
man’s arm, and its girth about that of his thigh. Its back fin, from
about one quarter of its length up to its head, is just like a brush, and
it has a liquid about it, sticky like gum, and when it fastens onto a
fish from below with this brush on its head the fish cannot get away, but
is held fast. On account of this peculiarity of the Hamby, the people
of Sambirano use it to fish with. When they catch one they confine it
in a cage of light wood, which they fasten in the sea, and.feed the fish
daily with cooked rice, or cassava, or small fish; and when they want to
use it, they tie a long string round its tail and let it go, following it in a
canoe. When it fastens on a fish they pull it in and secure the spoil.
There is a sea-turtle called by the people Fanéhana,® which the Hamby
is fond of catching, and this the people prize on — of the shell,
which is of commercial value.

The above account is given almost word for word by
James Sibree in his book ‘‘ A Naturalist in Madagascar,”
1915. Sibree, whose experiences in Madagascar cover a
period of fifty years, and who as his book shows was a
very close observer, evidently believed in this use of the

sh.

Tuer Huntine-FisH or THE West INDIES

However, the accounts quoted of the remarkable use
of the Remora as a hunting fish in the Mozambique coun-
try are not the first that we have of such employment.
For the very beginning we must go back to the second

5 This is probably the tortoise-shell turtle.

296 THE AMERICAN NATURALIST (Vou. LIII

voyage of Columbus to the New World in 1494. This
account given below is to be found in the writings of
Peter Martyr d’Anghera, who was a prominent figure
at the court of Ferdinand and Isabella and the foremost
letter writer of his day. In 1511 Martyr published at
Seville nine books and part of the tenth of his Decade I,
the Decade of the Ocean, one of the component parts of
his ‘‘De Orbe Novo,’’ which has since appeared in many
editions and translations. Possibly the best translation
available for the general reader is MacNutt’s, published
by Putnam in 1912, but as better preserving the spirit of
the times, I pricy to give Richard Eden’s translation
made in 1555, the quaint English of which reads as fol-
lows:

At the Ides of Maye, the watche men lokinge owte of the toppe
castell of the shyppe towarde the Southe, sawe a multitude of Tlandes
standinge thick together, beynge all well replenished with trees, grasse,
and herbes, and wel inhabyted. In the shore of the continent, he
[Columbus] chauneed into a nauigable ryver whose water was soo
hotte, that no man myght endure to abyde his hande therein any tyme.
The daye followinge, espyinge a farre off a canoa of fyshermen of th(e)
inhabitants, fearinge least they shulde flye at the syght of owre men,
he commaunded certyne to assayle them pryuily with the shyppe boates.
But they fearinge nothinge, taryed the comminge of owre men. Nowe
shal you heare a newe kind of fyshinge. Lyke as we with greyhoundes
doo hunt hares, in the playne fieldes so doo they as it were with a
huntyng fysshe, take other fysshes. This fysshe was of shape or
fourme vnknowen vnto vs: but the body thereof, not muche vnlyke a
greate yele: havinge on the hynder parte of the heade, a very towgh
skynne, lyke vnto a greate bagge or purse. This fysshe is tyed by the
syde of the boate with a corde litte downe soo farre into the water,
that the fysshe maye lye close hyd by the keele or bottom of the same,
for shee may in no ease abyde the sight of the ayer. Thus when they
espie any greate fysshe, or tortoyse (whereof there is great abundance
bygger then great targettes) they let the corde at lengthe. But when
she feeleth her selfe loosed, she enuadeth the fysshe or tortoyse as
swiftly as an arrowe. And where she hath once fastened her howld she
casteth the purse of skynne whereof we spoke before; And by drawyng
the same togyther, so graspeleth her pray, that no mans strength is
sufficient to vnloose the same, excepte by lyttle and lyttle drawinge the
lyne, shee bee lyfted sumwhat above the brymme of the water. For
then, as sone as she seeth the brightness of the ayer, she lettethe goo

No. 627] STUDIES IN ECHENEIS OR REMORA 297

her howlde. The praye therefore, beinge nowe drawen nere to the
brymme of the water, there leapeth soodenly owte of the boate into the
sea soo manye fysshers, as maye suffice to holde faste the praye, vntyll
the reste of the coompany haue taken it into the boate. Which thinge
doone, they loose so muche of the cord, that the hunting fysshe, may
ageyne returne to her place within the water: where by an other corde,
they let downe to her a piece of the praye, as we use to rewarde grey-
houndes after they have kylled theyr game. This fysshe, they caule
Guaicanum, but owre men caule it Reuersum. They gave owre men
foure tortoyses taken by this meanes: And those of such byggenes that
they almoste fylled theyr fysshinge boate. For these fysshes are
esteemed amonge them for delicate meate. Owre men recompensed
them ageyne with other rewardes, and soo let them departe.®

Curiously enough a repetition of this story by Martyr.
himself has been completely overlooked by all who have
had oceasion to refer to his Reversus story. I myself
did. not find it until, some two years after making notes
and copying his account as quoted above from Eden, I
chanced to go over the ‘‘Decades’’ again page by page
and stumbled on it. Since Martyr himself has not been
quoted directly it will be of interest to give this second
account from MacNutt’s excellent translation of Decade
VIII, Book 8, pages 299-300.

Let us now consider the hunting fish. This fish formerly vexed me
somewhat. In my first Decades, addressed to Cardinal Ascanio, I stated
amongst other marvels, if I remember properly, that the natives had a
fish which was trained to hunt other fish just as we use quadrupeds for
hunting other quadrupeds, or birds for hunting other birds. So are the
natives accustomed to catch fish by means of other fish. Many people,
given to detraction, ridiculed me at Rome in the time of Pope Leo for
citing this and other facts. It was only when Giovanni Rufo di Forli,
Archbishop of Cosenza, who was informed of all I wrote, returned to
Rome after fourteen years’ absence as legate of Popes Julius and Leo
in Spain, stopped the mouths of many mockers, and restored me my
reputation for veroai In the beginning I alab could hardly believe
the story, but I received my information from trustworthy men whom I

ers.
Everybody has assured me that they have seen fishermen use this fish
just as commonly as we chase hares with French dogs, or pursue the
wild deer with Molossians. They say that this fish makes good eating.
® This is a literal copy of Arber’s literal copy of Eden, save that the
old-fashioned {-shaped s has had to be replaced by the modern letter

298 THE AMERICAN NATURALIST [Vou. LII

It is shaped like an eel, and is no larger. It attacks fish larger than
itself, or turtles larger than a shield; it resembles a weasel seizing a
pigeon or still larger animal by its throat, and never leaving go until
it is dead. Fishermen tie this fish to the side of their barque, holding
it with a slender cord. The fish lies at the bottom of the barque, for it
` must not be exposed to the bright sun, from which it shrinks.

The most extraordinary thing is that it has at the back of its head a

sort of very tough pocket. As soon as the fisherman sees any fish

hi little cord. Like a dog freed from its leash, the fish descends on its
prey and turning its head throws the skin pouch over the neck of the
victim, if it is a large fish. On the contrary, if it is a turtle, the fish
attaches itself to the place where the turtle protrudes from its shell, and
never lets go till the fisherman pulls it with the little cord to the side of
the barque. If a large fish has been caught (and the fishermen do not
trouble about the small ones), the fishermen fasten stout cords to it and
' pull it into the air, and at that moment the hunting-fish lets go of its
prey. If, on the contrary, a turtle has been caught, the fishermen
spring into the sea and raise the animal on their shoulders to within
reach of their companions. When the prey is in the barque, the hunt-
ing-fish returns to its place and never moves, save when they give it a
piece of the animal, just as one gives a bit of quail to a faleon: or until
they turn it loose after another fish. I have elsewhere spoken at
length concerning the method of training it.” The Spaniards call this
fish Reverso, meaning one who turns round, because it is when turning
that it attacks and seizes the prey with its pocket-shaped skin.

This remarkable story of Martyr’s has been repeated
by many writers from his day almost to this and espe-
cially by the Spanish chroniclers of the early political
and natural history of the West Indies. Many of these,
however, add to the original story certain details which
will be of interest to include herein.

The first of these is the historian Oviedo, whose
‘*‘Sumario’’ was published but five years (1516) after
Martyr’s ‘‘Decades of the Ocean,’’ and whose ‘‘Chron-
icles’’ were first published in 1535. My excerpt is taken
from the Salamanca edition of 1547, but there is no rea-
son to think that this particular account differs from
that found in the earlier editions. We will let Oviedo

7 This account does not seem to have been preserved. At any rate it is
not to be found in MacNutt’s translation.

No. 627]. STUDIES IN ECHENEIS OR REMORA 299

speak for himself, and his account is’ all the more inter-
esting and valuable because he gives certain details as to
the training and care of the fisherman fish which are
absent from the other accounts, and of which he seems
possibly to have had some personal knowledge.

There is a fishing of these Manati and of the tortoise in the islands of
Jamaica and Cuba, which, if what I shall now say were not so public
and well known, and if I had not heard it from persons of great reli-
ability, I should not dare to write. And also it is believed that when
there were many Indians, natives, on the island Espagnola, they also
caught these animals with the Reversus fish. And since the discussion
of the history has brought me to speak of the animal, the Manati, it is
better that it is to be known that there are some fish as long or longer
than a palma, which they call the Reversus fish, ugly in appearance but
of great spirit and intelligence, which sometimes happens to be caught
in their nets along with other fish. This is a great fish and among the
best in the sea for eating, because it is dry and firm and without watery
parts, or at least it has very few; and many times I have eaten of it
and so am able to testify of it.

When the Indians wish to tame and keep any of these Reversus
fishes for their use in fishing, they catch it small and keep it always in
salt water from the sea, and there they give it food and make it tame,
until it is of the size which I have said or a little more, and fit for their
fishing. Then they take it out to sea in the canoe or boat, and keep it
there in salt water and fasten to it a cord delicate but strong. Then
when there is seen a tortoise or any of the large fish which abound in
these seas, or some of these Manati or whatever it may be that happens
to go on the surface of the water in such a way as to attract attention,
the Indian takes this Reversus fish in his hand and strokes it with the
other, and tells it to be manicato, which means strong and of good
courage and to be diligent, and other words exhorting it to bravery, and
to see to it that it dare to grapple with the largest and best fish that it
may find there [where the fishing is to take place]. And when the
Indian sees that the best time has arrived, he lets it go and even throws
it in the direction of the large fish. Then the Reversus goes like an
arrow and fastens itself on the side of a turtle, or on the belly, or
wherever it can, and thus clings to it or to some other large fish. This
one, when it feels itself seized by the little Reversus, flees through the
sea in one direction or another; and in the meantime the Indian fisher-
man lengthens the cord to its full length, which is many fathoms, and
at the end of this is fastenéd a stick or cork that it may be for a signal
or buoy which will remain on top of the water. In a little while the
` Manati or turtle, to which the Reversus has attached itself comes to the

300 THE AMERICAN NATURALIST [Vou. LII

shore, and then the Indian fisherman begins to draw the cord into his
canoe or boat and when there are but a few fathoms left, he commences
to draw it in carefully and slowly, guiding the Reverso and the prisoner
to which it is attached until they reach land and the waves of the sea
throw them out. The Indians who are engaged in the fishing leap out on
land and if the prisoner is a tortoise they turn it over even before it has
touched the ground and place it high and dry because they are great
swimmers; and if it is a manati they harpoon, wound and kill it. When
the fish has been taken to the land it is necessary very carefully and
slowly to release the Reverso which the Indians accomplish with soft
words, giving it many thanks for what it has effected, and thus they
release it from the other large fish which it captured and to which it is
so strongly attached that if it were forcibly removed it would be broken
to pieces; and thus in the manner I have described are taken these
large fish for whose chase and capture it seems that nature has made
the Reverso the sheriff and executioner. It has some scales similar to
the corrugations [grades] such as are found in the palate or upper jaw
of man or horse and therewith certain spines very thin, rough and
strong, whereby it attaches itself to the fish it seeks. And the Reverso
has these scales or corrugations full of these spines over the greater
part of the outer body, especially from the head to the middle of the
body along the back and not on the belly, but from the middle of the
body up; and from this circumstance they call it the Reverso because
with its shoulders it seizes, and fixes itself to fishes.

So eredulous is this generation of those Indians that they. believe the
Reverso well understands human speech and all those words of encour-
agement the Indian says before releasing it for an attack on the tor-
toise, manati or other fish, and that it understands also the thanks they
afterward give it for what it has done. This ignorance arises from a
failure to comprehend that this is a natural characteristic, because it
happens many times in the great ocean as I have frequently witnessed,
that when a shark or tortoise is captured, Reversos, without having been
directed, are found attached to these fish and are broken to pieces on
detaching them. From which we may infer that it is not in their power
to release themselves after they have attached themselves except after an
interval of time or from some other cause I have not determined; be-
cause one must think that when the shark or tortoise is taken the
Reversos attached thereto would flee if they could. The fact is, as I
have said above, for each animal there is its constable.

In 1527, Benedetto Bordoni published his ‘‘Isolario.”’
In it is a brief account of the fishing in that locality called
Queen’s Gardens. It seems to be an abbreviated tran-
seript from Peter Martyr and adds nothing new, save a

No. 627] . STUDIES IN ECHENEIS OR REMORA 301

map of Cuba, showing the islands off the southern coast
among some of which the fishing, with the Guiacan was
observed. This seems to be of enough interest to be re-
produced herein as text-figure 3.

In 1553, Gomara published at Medina del Campo his
‘‘Historia General de las Indias.” On folio XIIIT is
found an abbreviated copy of Oviedo’s account of the
Reversus fish, but as it contains nothing new it need not
detain us.

S Trey
oe

A

TEXT-FIGURÐ 3. The Island of Cuba with the Jardinellas de la Reina to the south.
(After Bordoni, 1527.)

The greatest of the encyclopedic writers on natural
history in the Renaissance times was the Swiss, Konrad
Gesner, who was too good a searcher for the marvellous
to let such a story as this escape him. His account (1558)
is a somewhat abridged but yet almost literal translation
of Peter Martyr. However, he gives us a figure of a
hunting scene, showing how this fisherman-fish was used,
and this is reproduced herein as Fig. 4, Plate I. The
Reversus fish is shaped like an eel and has a great bag or
pouch attached on the back of its neck. This pouch has
just been thrown over the head of what appears to be a

seal (probably meant for a manatee), while a turtle looks
on in amazement from one side. Inthe background in this

PLATH I

Fic. 1. Soned disk of Remora. “After Jordan and Evermann, 1906.
Fie. 2. Leptech

e rda Eve

Fis. 3 reai brachyptera. After Jordan and Eve

Fic. 4. The first known figure of fishing with the fistierman fish, After Ges-
ner, 1558.

No. 627] STUDIES IN ECHENEIS OR REMORA 303

boat are the fishermen, one of whom holds one end of a
line the other end of which is tied around the anterior
part of the body of the eel-like fish. In a sort of post-
script Gesner refers to another hunting-fish which is
similar to but smaller than the above. This reference,
however, is not clear.

The first user of the name Guiacan for our fish was
Peter Martyr; other and later writers take the name
from him. Considerable effort was made to run down
this word and to ascertain its meaning. It was finally
found in Bachiller y Morales’s ‘‘ Cuba Primitiva’’ ie
Here we are told that

Guiacan was the name the Indians gave to the fish which the Span-
iards called Reverso, and which served them in fishing; because tied by
the tail, they fixed themselves to the tortoise and other prey which they
did not release, rendering thus a useful service.

Earlier than Bachiller y Morales, another writer, Ray-
mond Breton (1665), calls the huntsmen fish ‘‘Iliouali’’
and says that it is a fish which has on its head a mem-
branous plaque, and if it attaches itself to the canoe it
can with difficulty be removed save by breaking it into
fragments.

That part of Gesner’s ‘‘Natural History of Animals’’
which has to do with fishes was worked over in German
and published in 1575 as ‘‘Das Fischbuch.’’ In it on
page L is found the figure of the hunting scene just re-
ferred to and an abbreviated account of the use of the fish
as a living fish-hook. Here also there is an account of

8 Every effort has been made to ascertain the original of this figure.
Presumably it is from an insert in some contemporary map or similar pub-
lication, Dr, Eastman personally made a search through the rich collection
of Americana in the New York Publie Library, the able curator of which,
Mr. V. G. Paltsits, had to confess himself at a loss. I myself have worked
through the collection of reproductions of old maps in the same library but
in vain. Finally the question was submitted to Mr. E. A. Reeves, the
learned curator of maps of the Royal Geographical Society, London, who
courteously made a lengthy search through all the old maps under his care
Finding nothing he passed the question along to the authofities of the
British Museum, who in turn could give no co So the origin of this

interesting and oldest figure still remains a mystery

304 THE AMERICAN NATURALIST [Vou. LIII

another Reversus. Apparently herein Gesner has mixed
certain data from Oviedo with the legends of another
Reversus covered with sharp spines.

It seems that in the writings of these old Spanish his-
torians two fishes are described called Reversus ;? one the
anguilliform kind, having a pouch or sucker on its head,
evidently a Remora, or, since it grows larger, an Eche-
neis; the other the squamous kind covered with scales
bearing long spines, evidently the swell fish, Diodon.
Concerning these fishes Dr. C. R. Eastman has written
several interesting and valuable papers to which the at-
tention of the reader is called. (See Bibliography, East-
man 1915, 1915a, 1916.)

We next hear of the Reversus in the writings of one
Antonio Galvano. His book, ‘*The Discoveries of the
World from their first Original unto the Yeare of our
Lord 1555,’’ was published in the original Portuguese in
1563 under the editorship of his friend, F. Y. Sousa
Tavares, and translated and reprinted at London in 1601
by Richard Hakluyt. Neither of these editions being
available. Ihave had to content myself with the Hakluyt
Society’s reprint'® found in Vol. 30, 1862, as edited by
C. R. D. Bethune. Here there is a short paragraph in
which the use of the anguilliform eel is attributed to the
squamous form. Nothing new is added and no quotation
will be given.

9 The reversus or ‘‘upside down’’ fish was undoubtedly so named because
when attached to the carapace of a turtle its belly was turned upward or
outward, as also when it was attached to the side of a fish—in any case its
natural position was reversed. Diodon when it yaa its belly with air
floats at the surface belly up, hence it e was a Reve fish.

10 Tt is interesting to note that in the Hakluyt an i se Reverso story is
put in square brackets. This considerably confused me and lest others be
similarly thrown off the track it seems well to add this note from Mr. C. K.
Jones of the Library of Congress, face agi che publishing his 1601 edi-
tion was unable to find the original. iety in preparing its
1862 edition secured a copy of the Gena. eee SE. of 1563 from John
Carter Brown; and from this copy the Portuguese text was printed.’’ It
seems that Hakluyt included in his 1601 edition the Reverso story from
original histories, However, in the original Portuguese text, Mr. Jones finds
the Reverso story without brackets.

No. 627] STUDIES IN ECHENEIS OR REMORA 305

We next hear of the fisherman-fish in Herrera’s ‘‘ His-
toria Generale de las Indias Occidentales’’ published in
1601. In Capt. John Stevens’s translation we read:
` They [the Indians] fished on, and took some fishes they called reves,™
the biggest of them about the size of a Pilchard, having a roughness on
the belly [?], with which they cling so fast, wheresoever they first take
hold, that they must be torn in pieces before they can be torn off again.
They ty’d these by the Tail with a small Thread, about two hundred
Fathoms more or less in Length, and the Fish swimming away on the
Surface of the Water, or but a little under it, when it came to where
the Tortoise was in the Water, it clung to the under Shell thereof, and
then the Indians drawing the thread, took a Tortoise that would weigh
a hundred Weight, or upwards. After the same manner they took
Sharks, which are most cruel bloody Fishes that devour Men.

Next comes Ramusio, whose ‘‘Della Historia dell’
Indie’’ bears date Venetia, 1606. This appears to be
merely a translation into Italian of Oviedo’s Spanish
work. At any rate it adds nothing-to our knowledge of
the hunting-fish, and may be passed over with this brief
notice.

Another of the ‘‘fathers’’ of ichthyology is Aldro-
vandi, whose great work was published in 1613. He
figures and describes both kinds of the Reversus. In
general he follows Peter Martyr, but it is very clear that
he copies Gesner. However, he has had Gesner’s fishing
scene redrawn, as may be seen from the reproduction of
it herein (Fig. 5, Plate II). The boat and boatman are
omitted, as is the cord around the neck of the fish, the
seal-like animal has been replaced by another probably
intended to represent a manatee, the turtle is entirely
different, and lastly the head of the Reversus is not at all
that of Gesner’s figure. This is much larger, the teeth
are more marked, the upper jaw has a hooked beak; and
the bag of skin comes more distinctly off the top of the
head, and is smaller at the base and has more longitu-
dinal striations. And yet for all these changes it is
plainly Gesner’s figure.

11 Reves is of course a variant of the word Reversus, an abbreviation
possibly,

daah arig
y ? e

E ie 73 8

PLATE II
Fic. 5. The Indian anguilliform Reversus. After Aldrovandi, 1613.
Fic. 6. Reversus or Guiacanus, according to Nieremberg, $
Fic. 7. Fishing with the Reversus, from Ogelby’s “ America,” 1671.

No. 627] STUDIES IN ECHENEIS OR REMORA 307

In another place Aldrovandi gives a figure of the
spinous Reversus, but in his account of this form he gets
his data badly mixed since much of it is the data which
Peter Martyr ascribes to the anguilliform variety. In
neither account does Aldrovandi offer anything new.

We now come to a Spanish work published in Mexico
City five years before the Pilgrims landed on Plymouth
Rock and when Jamestown was but eight years old. This
is Hernandez’s work (1615) on the nature and virtues of
the plants and animals used in the practise of medicine
in New Spain. How he brings in the Remora is not
clear, but he attributes his account to Oviedo, the actions
of whose anguilliform Reversus he describes in his
(Oviedo’s) own words. However when he attempts to
further describe the fish he gets his account tangled up
with that of the porcupine fish. He does not seem to
‘have ever seen either fish.

In 1635, Joannes Eusebius Nieremberg, a Jesuit priest,
who was professor of physiology in the Royal Academy
of Madrid, published his ‘‘Historia Naturae’’ in folio
form. This is a compilation of not very great value, the
less so because the references are not set forth clearly.
Our interest in his book, in which he quotes Peter Martyr,
Oviedo, Hernandez and another to be referred to later, is
chiefly centered in his figure of the Reversus or Guiaca-
nus. This is reproduced here as Fig. 6, Plate II. This
is plainly Gesner’s figure with the addition of a sort of
saw-toothed mane on the anterior dorsal region.

Ogilby, whose huge tome was published in 1671, had
evidently never seen the Guiacan, but he inserted on page
49 of his ‘‘America’’ such a quaint and interesting figure
of his conception (or his artist’s) of how this fishing was
carried on, that this is reproduced herein as Fig. 7,
Plate IT.

The Dutchman, Th. van Brussel, in 1799 published a
very interesting account of the Reversus; but a careful
translation of his Dutch shows that it is but a translation
of Martyr and Oviedo, and further that he confuses the

308 THE AMERICAN NATURALIST [Vou. LIII

anguilliform and squamous forms of the Reversus—a
figure of the latter being given. He also need not de-
tain us.

From this time on a long succession of writers repeat
the tale. Thus we find it in Shaw’s ‘‘Zoology,’’ Vol. IV,
1803; Humboldt’s ‘‘Essai Politique sur 1’Ile de Cuba”’
(1826), his ‘‘Receuil d’Observations de Zoologie et Ana-
tomie Comparée (1833) and in the ‘‘Personal Narra-
tive’’ (English translation, 1860). We also find it in
most if not all of the ‘‘Lives’’ of Columbus, notably
Irving’s (1828), Winsor’s (1892), and last and best
Thacher’s (1903).

To these foregoing accounts we may add a brief note
which may be of interest. Bernabe Cobo was a Spaniard
(born 1582, died 1657) who wrote his ‘‘ Historia del Nuevo
Mundo” and at his death left it in manuscript where it
remained until found, edited and published by the Span-
ish naturalist, Mareos Jimenez de la Espada, towards
the close of the last century. Volume II, Sevilla, 1891,
contains Cobo’s story which turns out to be the familiar
paraphrase of Oviedo’s account. Absolutely nothing
new is added.

We now come to a consideration of the sources of the
various accounts of the use of the sucking fish as a living
fish-hook in the West Indies. First of all plainly these
later accounts are all echoes of Peter Martyr, or of
Oviedo, or of both. Then these further questions nat-
urally arise: Is Peter Martyr’s ‘‘Decade of the Ocean”’
in 1511 the first account published? And secondly what
is the ultimate source of these earliest accounts? In
answering these questions I have had three invaluable
sources of information. The one is Justin Winsor’s
keenly critical life of Christopher Columbus, the second
is John Boyd Thacher’s monumental work on Columbus
(Vol. II, 1903) and the third is the continued advice and
unfailing help of my friend, the late Dr. Charles R. East-

No. 627] STUDIES IN ECHENEIS OR REMORA 309

man.'* Dr. Eastman became interested in the subject
while working on the great ‘‘Bibliography of Fishes”
published by the American Museum of Natural History,
and finding that I was collecting data for a series of
papers on Echeneis most courteously turned over to me
invaluable material and aided me in every possible way.
At the very time when I was slowly tracing these accounts
backward towards their ultimate source, Dr. Eastman in
the most brilliant fashion ran these stories down to the
original recorder himself.

First of all let us see if Martyr’s account in 1511 is the
first published account of the interesting phenomenon.
To this the answer must be ‘‘No!’’ Dr. Eastman sent
me the following extract from ‘‘Libretto de Tutta la
Navigatione de Re de Spagna et de le Isole et Terreni
Novamente Trovati,” Venezia, April, 1504 [‘‘A Little
Book in Regard to All the Navigation of the King of
Spain to the Islands and Newly Discovered Lands’’]:

Continuing [along the coast of Cuba] they found further onward
fishermen in certain of their boats of wood exeavated like zopoli, who
were fishing. In this manner they had a fish of a form unknown to us,
which has the body of an eel, and larger, and upon the head it has a
certain very tender skin which appears like a large purse. And this fish

[biscia] they loosen the noose, and this fish at once darts like an arrow
at the fish or reptile, throwing over them this skin which he has upon
his head; which he holds so firmly that they are not able to escape, and
he does not leave them if they are not taken from the water, but as soon
as he feels the air he leaves his prey and the fishermen quickly seize it.
And in the presence of our people they took four large turtles which
they gave our people for a very delicate food.

After Dr. Eastman had sent me the above translation
from the Libretto, I very carefully worked over Volume
II of John Boyd Thacher’s monumental life of Columbus

12 The recital may perhaps not be devoid of éither interest or value if
the steps are set forth by which Dr, Eastman and myself, working —
and at a great distance from each other, traced this interesting story back
to its original narrator. But it should be said here that man
reached the goal first, and that my efforts were chiefly confined to T
his results, and clearing up certain details.

310 THE AMERICAN NATURALIST (Vou. LIII

and from it much of the data following have begn ob-
tained. Only one copy of the Libretto is known in the
world, and it is preserved in the San Marco Library at
Venice. Thacher traced the original manuscript copy
of the Libretto to the ownership of a man named Sneyd,
living at Neweastle-on-Tyne, but was refused even the
sight of it much less a chance to make photographs.
However the authorities of the San Marco Library were
men of different caliber, and Thacher reproduces in his
book the whole Libretto page by page. And I in turn
reproduce here as text-figure 4 a part of Thacher’s re- |
production of the page giving the Reversus story. It is
from chapter XV.

a Ceri Trouarono dapo! piu auati al-
cuni pelcadori i certe fue barche de uno legno cauo come zopoli ch pe
{cauio.In to m5 haucuao tn pefce duna forma a noi incognita ch ha
el corpo đ aguiila:& mazor:& fupra ala tefta ha cerca pelle teneriffima
che par una borfa grade. Er q{to lo tiguno ligato cé una trezola ala fpo
da dela barcha p che el nō po patir uifta de aere:& côe uedeo alchun pe
fce grade o bifia fcudelera li laffao la trezola:& qllo fubito corre como
una {gets al pelce o ala bifcia:butidoliadoffo qlla pelle ch tien fopra la
tefta cô laĝil tié tato forte ch {cipar nô poffono:& non lı Jaffa fi nol tiri
for de lag: <lqi futiro fentiro laire lafla la preda.& li pefcadori pito api-
glare.Er i pňtia de li nři Hfero.iiii.gran caladre,leq'e donorong ali nti p
cibo dilicatiffimo.

TEXT-FIGURD 4. Page from the Libretto, 1504, whereon is contained the

first printed account of the fisherman fish. Reproduced from Thacker’s “ Chris-
topher Columbus,” II, 1903 :

The Libretto of 1504 was the first collection of voyages
to the new world ever printed, and as such is of great
interest to. scientific men for more reasons than those
merely pertaining to this article; hence it may be of in-
terest for us to consider for a few minutes its history,
which is as follows.

Peter Martyr, born in Italy, was a courtier and literary
man of high standing in the entourage of Ferdinand and
Isabella. Thacher says: ‘‘Peter Martyr d’Anghera may
be said to have composed the matter in this little book,
writing it in Latin from a series of letters addressed by

No. 627] STUDIES IN ECHENEIS OR REMORA 311

him to various noted persons. These letters were writ-
ten immediately after the events they describe. They
bear the first news. They reflect first impressions... .
This work was put into its present narrative form some
time prior to the summer of 1501.”
There now enters upon the scene another Italian letter
. writer, one Angelo Trivigiano, who was secretary to
Domenico Pisani, the Venetian ambassador at the Span-
ish court. Thacher publishes copies of three letters
which Trivigiano wrote in 1501 to the Venetian admiral
and historian Domenico Malipiero (whose retainer he
seems to have been) transmitting copies of various sec-
tions of a ‘‘voluminous work’’ on the voyage of Colum-
bus ‘‘composed by an able man.” Trivigiano nowhere
names Peter Martyr as the author, but in all three of the
letters he says that the author is the ambassador of the
Spanish court to the Sultan of Egypt, and contemporary
history informs us that this was no other than Peter
Martyr, who left Granada for Egypt, August 14, 1501.
The contents of the Libretto, in Peter Martyr’s own
words, baring an introductory paragraph by Trivigiano
descriptive of the personal appearance of Columbus, was
turned over by Malipiero to Albertino Vercelles da
Lisona, and by him issued in the Venetian dialect as a
printed book on April 10, 1504.78
(To be Continued)
13 The only other historian of Columbus whom I have found to make
mention of the Libretto is Winsor, who says that the first seven books of the

first Decade were sent in Italian to Venice and there issued as a printed
book of 16 leaves in April, 1504,

THE GERM PLASM OF THE OSTRICH

PROFESSOR J. E. DUERDEN
RHODES UNIVERSITY COLLEGE, GRAHAMSTOWN, SOUTH AFRICA
A

The germ plasm is fundamental and remarkably conservative .. .
when the germ plasm changes it does so as a result either of wholly
internal physiological causes, or of very extraordinary environmental
stresses acting directly upon the germ cells . ... mixing of germ plasms,
in and of itself, does not mutually alter hereditary determiners. . .
selection only acts as a mechanical sorter of existing diversities in the
germ plasm and not as a cause of alteration in it.

B

Hereditary determiners or factors fluctuate regularly and frequently,
if not indeed usually, and in high correlation with somatic characters

. mixing of germ plasms in fertilization alters hereditary deter-
miners mutually and hence is, in and of itself, a cause of genetice varia-
tions . . . a purely external agent, the continued selection of personal
somatic qualities, will alter the germ plasm.

In the above clear, concise phrases, sometimes with
supporting amplification, Dr. Raymond Pearl,’ in the
presidential address before the New York Meeting of the
-American Society of Naturalists, 1916, contrasts the atti-
tude of two sections of American geneticists with regard
to the manner of changes in the germ plasm, as affording
so much somatie material upon which selection may pos-
sibly work in the evolution of animals and plants.

o much evidence is already available for discussion on
the merits of the one side or the other that it would appear
gratuitous to add more, and one can well appreciate the
advice which Pearl gives to get down to more, and more
searching, investigations as to the causes of genetic (fac-
torial) variation. The case of ostrich breeding in South
Africa however affords such direct evidence bearing upon

1‘*The Selection Problem,’’ AMERICAN NATURALIST, February, 1917,
ol, 51.

312

No. 627] GERM. PLASM OF THE OSTRICH 313

most of the dicta that it is thought an account may be wel-
comed by geneticists. At any rate it may be added to the
already voluminous ‘‘ Experience of Practical Breeders,’’
containing facts which will need to be reckoned with in
any explanation of the actual causes of germinal changes.
The ostrich affords an example of an animal only recently
domesticated and still in the making, and we have before
us the practical methods followed and the results ob-
tained, enabling us to deduce in some measure the genetic
principles involved. The endeavor will be to see what
contribution its germ plasm has to make to each of the
contrasting statements at the head of the paper, not for-
getting that we know but little of the nature of the germ
plasm and its changes except from their manifestation in
the soma. It may be there is truth in both attitudes.

I

“The germ plasm is fundamental and remarkably con-
servative.’’

Ostrich farming on methodical lines was first under-
taken in South Africa about fifty years ago. The begin-
nings were made with chicks obtained from wild nests, as
unless ‘‘tamed’’ from an early age control of the adults is
afterwards impossible. So remunerative did the industry
prove to be that with the exception of one or two setbacks
it advanced with great rapidity until at its zenith, the year
before the war, 1913, nearly 1,000,000 domesticated birds
were recorded, yielding an export of 1,023,307 lbs. of
feathers at a value of $15,000,000, forming with gold and
diamonds a triad contributing much to the prosperity of
South Africa. With the advent and continuance of the
war depression of a most severe character set in among
ostrich farmers, and the number of birds has been re-
duced by about two thirds.

In the early days of the industry very little account was
taken of the quality of plumage produced, and any bird
reaching sexual maturity (three to four years) was em-
ployed as a breeder. Within the past two or three decades

314 THE AMERICAN NATURALIST [Vou. LII

however the greatest attention has been devoted to the
many characters of the plume and only the best plumage
birds have been employed as breeders, the chief reason
being the great difference in returns from clippings of
high quality compared with those of an ordinary or infe-
rior type. An intensive study has arisen in connection
with the various structural details of the feather and also
with the measures necessary for their production in the
highest state of excellence; among the latter are included
_ both the feeding and management of the birds as well as
selection in breeding. It is probably safe to say that no
domestic animal has been more intensively and intelli-
gently studied by the farmer than the high grade ostrich,
or more pampered in its treatment. Breeding sets, a cock
and a hen, known to produce progeny giving superior
plumage have frequently realized as much as $5,000.

The ‘‘points’’ of the ostrich plume relate to details con-
cerning the length, width, density, lustre, shapeliness and
evenness of the flue (vanes) and the form and strength of
the shaft, and a highly technical terminology has arisen
in connection therewith. An ostrich produces annually
from 200 to 300 commercial feathers, belonging to a dozen
or more different classes—whites, byocks, blacks, drabs,
floss, tails—each with its many subdivisions. Hach indi-
vidual feather is handled and specially examined several
times in the processes of clipping, arranging, sorting and
selling, before being exported, and prior to the war two or
three hundred millions of feathers were in this manner
passed in review.

Under such keenly discriminating circumstances it will
be understood that if any plumage variation presented
itself it would be at once recognized and brought to gen-
eral notice. A bird giving rise to a departure of any
moment in a desirable direction in connection with any
of the feather points mentioned would represent a fortune
to its owner. But not a single case has ever been forth-
coming. Without any hesitancy it can be affirmed that in
the course of the fifty years during which the ostrich has

No. 627] GERM PLASM OF THE OSTRICH 315

been domesticated it has never produced a feather varia-
tion, germinal in its origin, such as could be regarded as of
the nature of a sport or mutation. Feather irregularities
and abnormalities are by no means infrequent, but can
generally be ascribed to some injury to the feather germ
or follicle in the process of quilling, or to malnutrition.
Any peculiarity of this nature is usually forwarded to the -
writer, and some of the more common irregularities have
already been described? They are never hereditary
peculiarities.

This stability on the part of the various structural de-
tails of the feather has continued despite the great changes
to which the ostrich has been subject as a domesticated
creature. The birds are fed on the most nourishing and
stimulating of foods, the farmer having no option in the
matter if he is to secure a feather crop of the highest
quality; also they may be transferred from the moist
coastal planes to the dry and arid interior at an elevation
of 5,000 or 6,000 feet, a change involving great variation
in temperature, pressure and other conditions. As a
epidermal product, growing at the rapid rate of a Guari
of an inch daily, the feather is extremely sensitive to
changes in nutrition and climatic conditions, often re-
sponding to the small differences in blood-pressure be-
tween day and night. Yet all the modifications resulting
from these influences are somatic; no hereditary germina]
alteration has ever manifested itself.

Like so many other African animals, the giraffe, hippo,
rhino, elephant and ant-bear, the ostrich is a survival of
ancient days, a left-over, and as becomes a creature of
long ancestry is fixed and immutable with regard to the
many characteristics of its plumage. Numerous germinal
changes have appeared in the past and survive to-day in
the various feather types recognized by the specialist, all
of which breed true; but it can justly be claimed that no
further alteration has taken place during the past fifty

2**Experiments with Ostriches, ee Sone Irregularities,’’ Agric.
Journ., Union of South Africa, August, 1

316 THE AMERICAN NATURALIST (Vou LITI

years, in spite of the many environmental changes to
which the bird has been subject. As regards the struc-
tural details of the feather the germ plasm of the ostrich
fully confirms the statement with which the section opens.

H

‘Mixing of germ plasm, in and of itself, does not mu-
tually alter hereditary determiners.”

If the plumage characters of the ostrich are so immu-
table what then is the objective in breeding? The original
wild stocks with which the farmer commenced in the
sixties differed.much among themselves in the structural
minutiæ of the feather, and the most desirable of the
various feather points were distributed among many
strains. The earnest endeavor of the ostrich breeder is
to combine in the single plume the best of all the many
desirable features originally scattered throughout the
wild birds. The ultimate purpose of every breeder is the
same—to produce a plume combining the maxima of all
the available feather characters; a plume having the
greatest length, width, density and luster and the most
perfect shape, supported on a round, strong, slender
shaft. On the original birds the largest plumes had for
the most part a coarse, loose, unshapely flue, while the
most compact, shapely, lustrous, graceful plumes were
generally small. The whole effort is to combine the maxi-
mum size with all the so-called ‘‘quality points’’; no other
feature of the bird is taken into account in breeding, as
none has any commercial value or is known to be in any
way correlated with feather production. The problem
appears simple, though it is taking years to accomplish;
progress is being made each year, but the ideally perfect
ostrich plume is not yet.

-= The genetical methods of the farmer are likewise
simple. He proceeds entirely in the belief of a blending
inheritance, which though doubtful in theory is succeed-
ing in practise. He starts with a bird which produces
plumes the most nearly approaching his ideal, and mates

è
No. 627] GERM PLASM OF THE OSTRICH 317

it with another most closely resembling it, but perhaps
lacking or surpassing in one or more points; another sea-
son he may resort to a different mating to secure other
features. From different breeding sets he may rear two
or three hundred chicks in a season. The progeny being
mostly intermediates and showing much variation he
selects when mature the most desirable among them as
breeders, or maybe, being weak in some particular point,
he will purchase or exchange with another breeder in
whose birds the character is strong. By this method,
essentially one of hybridization, the ideal plume is being
slowly built up. Sometimes by a fortunate mating one
breeder will be ahead and sometimes another, a success-
ful competitor at a Feather Show being inundated with
orders for breeding birds and chicks and his fortune well
assured. Despite the variability in the progeny no
breeder can afford to ‘‘fix’’ his strain by a measure of
inbreeding, lest while doing this another may get ahead.
Taking all the economic and biological circumstances
into account the geneticist has little he can contribute to
such a practical effort; he can but assist by endeavoring
to deduce and explain the principles involved.
The textual application is manifest. The greatest mia-
_ture of germ plasm is going on, but no single hereditary
factor or determiner is altered in the process, and has not
altered throughout the history of ostrich breeding; only
new combinations are formed of factors already available.
The farmer himself has long grasped this and does not
look for any change; he knows he can get nothing beyond
what the wild bird had to start with; he can create or
change nothing, beyond what can be ascribed to good
management and feeding. For practical purposes his
understanding of the individuality and fixity of the germ
factors producing the plume is as clear as that of the
most zealous Mendelian; but only in a few instances has
he ever heard of the factorial hypothesis, though facts
upon which it could have been established were discovered
in his farming practise long before 1900, the year of Men-

318 THE *AMERICAN NATURALIST [Von. LIII

delian reawakening. If his birds, judging by their feather
performance, are lacking a certain germ factor he is well
aware that he can by no possible means originate the
factor nor hope to produce it in any way; he must pro-
cure it from some other farmer whose birds display it,
and then he may expect to secure it in combination in his
own strain.

Hil

“Selection only acts as a mechanical sorter of existing
diversities in the germ plasm and not as a cause of altera-
tion in it.”

The term ‘‘selection’’ is employed by the ostrich
breeder in South Africa with all that freedom which Pearl
finds among the plant and animal breeders in America,
but he is never under any delusion that it signifies more
than is implied in the simple meaning of the word. He
has retained its plain everyday significance and the major-
ity have never heard of Darwin and ‘‘The Origin of
Species by Means of Natural Selection,’’ nor of the ex-
tended meaning which students of evolution are inclined
to give the term, as in the phrase, ‘‘The Selection Prob-
lem.” To the ostrich farmer ‘‘breeding from selection’?
simply means that for his breeding sets he picks out birds
having the special plumage characters he desires to see in
their progeny, or which he expects to get from the com-
bination of the cock and the hen. Selection is merely used
in contrast with indiscriminate breeding, as,where any
cock and hen may be camped off without regard to their
plumage value, or in contrast to breeding on the veld
where any cock may mate with any hen. He selects partly
on the basis of somatic performance and partly on proved
germinal production; many birds which themselves give
indifferent plumage are yet employed as breeders from
being known to produce superior chicks.

From his life-long experience in selective breeding the
ostrich farmer clearly grasps that all he is doing is to sort
out birds from among his flock with certain characters

i

No. 627] GERM PLASM OF THE OSTRICH 319

which he desires to see in combination in their progeny,
' but he never dreams that any change in the characters
themselves will result therefrom. Though perhaps un-
able to express it in words he knows that the germ plasm
of each of his birds contains so many factors, and in his
selection of breeders picks out the birds having the
factors he desires to give him new combinations, but he
has no expectancy that the factors themselves will
undergo any change as evidenced by their expression in
the progeny. Selection along the prescribed lines is prob-
ably as rigid as that which any experimentalist could
carry out, and is certainly more so than can be conceived
of as taking place in nature, yet long as it has been in
operation it has never carried with it an alteration of any
of the existing diversities of plumage.

IV

“When the germ plasm changes it does so as a result
either of wholly internal physiological causes, or of very
extraordinary environmental stresses acting directly upon
the germ cells.

The bodily characters of the South African ostrich pre-
sent a remarkable uniformity except as regards certain
details to be described later, but in comparison with the
North African bird many striking differences appear. In
1912 the Government of the Union of South Africa im-
ported 132 specimens of the northern ostrich from
Nigeria. It was hoped that in these some one or other of °
the plumage characteristics might be developed to a
higher degree than in the southern bird and could with
advantage be combined with the latter. Experiments
with this end in view are now in progress under the direc-
tion of the writer.

The northern ostrich is longer in the legs and neck than
the southern, the head reaching a little over eight feet
from the ground, about a foot more than in the latter.
The color of the skin of immature birds of both sexes and
of mature hens is a creamy yellow, while the mature cock

320 THE AMERICAN NATURALIST (Von. LII

is bright red or scarlet on the legs, head and neck, and red
and pink over the body generally; in the southern ostrich
the skin of the neck, body and legs is a pale yellow in
chicks, dark gray in mature hens and dark blue in cocks,
while in the sexually ripe cock only the beak, the front
part of the head, the naked skin around the eyes and the
tarsal scales are a bright scarlet. The crown of the head
of the northern bird has a bald oval patch while that of
the southern is covered with hair-like feathers similar to
those over the rest of the head and neck. The northern -
egg is larger and rounder, with an enamel-like smooth-
ness, and is practically free from obvious pittings; the
southern is deeply pitted all over, smaller and more oval.
Knowing as we do the habits and life of the ostrich it is in
the highest degree improbable that any of the differences
have an adaptive significance or selective value in nature.

When the birds are observed side by side, as can now be
done at Grootfontein, the above characters readily serve
to separate the northern from the southern ostrich, and
may well be held to justify the specific distinction usually
accorded them. That the distinguishing features of the
former are not environmental but germinal is proved by
the fact that they persist under southern conditions and
have reappeared in progeny already reared. Numerous
cross-breds or hybrids of the first generation have also
been obtained, but sufficient time has not yet intervened
to secure the second hybrid generation. As regards di-
mensions, color and the nature of the egg the first gen-
eration of cross-breds are intermediates in varying degree
between the northern and southern parents, but the bald
head patch of the northern is dominant over its absence
in the southern, appearing in all the crosses yet reared.

It is clear that the germ plasm of the northern ostrich
has undergone marked changes compared with that of its
southern representative, or vice versa, for one can only
think of the various races or species of Struthio as derived
from a common stock. In terms of the sectional heading
we may well enquire whether the changes are due to

No. 627] GERM PLASM OF THE OSTRICH 321

internal physiological causes or to extraordinary environ-
mental stresses acting upon the germ cells. We have
already seen that germinally the South African ostrich is
most irresponsive to any environmental changes and we
have no reason to suspect that its northern relative is in
any way more impressionable. In the climatic and other
environmental conditions of North Africa it is difficult to
conceive of anything which could, for example, modify the
bodily colors as compared with those of the southern bird,
or could bring about a perfectly smooth round egg in con-
trast with an oval pitted one, much less which could either
directly or through the soma change the germ cells so as
to render the differences hereditary. Of course we know
next to nothing of the influence upon the germ cells of
extraordinary environmental stresses and to labor the
point would be unprofitable. But doubt may certainly be
expressed as to whether any external influence could so
change them as to bring about the formation of a bald
head patch, a feature which it is impossible to regard as
having an adaptive significance. It is a new germinal
character which has appeared in the northern bird, en-
tirely sui generis; there is nothing suggestive of it in the
southern ostrich.

We have the hard fact to account for that the germ
plasms of the northern and the southern ostrich differ
from one another in certain respects as revealed by their
manifestation in the soma, and it is also proved that they
breed true irrespective of environment. And while in our
condition of absolute ignorance no good purpose will be
served by dogmatizing it may be permitted to express the
conviction that the germ plasm changes as between the
northern and southern ostrich have resulted entirely from
internal physiological causes. The conviction is strength-
ened all the more from the facts to be presented in the
next section.

322 THE AMERICAN NATURALIST (Von. LII

B?
y

“Hereditary determiners or factors fluctuate regularly
and frequently, if not indeed uputu and in high correla-
tion with somatic characters.’

In a certain measure this statement may be looked upon
as opposed to that with which section 1 opens, but no one
would maintain either the one or the other to be the ex-
clusive state of the germ plasm, hereditary determiners
or factors of all animals and plants. We have abundant
evidence that the germ plasm is remarkably conservative
for some forms of life (persistent types) while in others
it may fluctuate or change frequently (Drosophila) ; also,
it is not unreasonable to expect that at any one period the
factors for certain parts of an organism may remain fixed
while for others they may be in a state of change. We
have seen that the factors controlling the structural de-
tails of the ostrich plume are peculiarly constant, but the
endeavor will now be made to establish that those for the
wing feathers numerically, as well as for certain other
parts of the bird, are undergoing regular and frequent
changes and in determinable directions.

By zoologists the wing of the ostrich is usually regarded
as degenerate, on account of its small size compared with
the body and legs and the practical absence of any cover-
ing of feathers on its inner or under surface. Certain
studies recently made have given good reason for con-
cluding that in many other less obvious respects it is still
undergoing degeneration. The full details upon which

8 The three statements under section B are obviously considered by Dr.
Pearl to apply specially to Dr. Castle’s claims in connection with his ex-
periments on piebald rats, a condensed account of which appears in the same
issue of the NATURALIST as Pearl’s paper (p. 102). Instead of regarding
them as applicable only to the disputed plus and minus fluctuations in the
factor itself it may be permissible to consider them in a broader sense, as
referring to the nature of the germ-plasm generally and as contrasted with

ose in section A. What follows has probably no connection with results
uch as those which Castle has obtained but nevertheless it is hoped to show
that real factorial changes, continuously retrogressive in their nature, are

going on in the germ plasm of the ostrich and that ne | is much likelihood
the changes ean be influenced by selection.

No. 627] GERM, PLASM OF THE OSTRICH 323

the claim is based will appear later. Only the outline of
the facts can now be given in so far as they bear upon
the condition of the germ plasm.

Only a single row of under-coverts usually occurs on
the wing of the ostrich, its members alternating with the
remiges or wing quills (Fig. 1). In but two specimens
out of hundreds examined however has the full number
of feathers required for alteration with the complete row
of remiges been found. Usually eight to ten are missing
from the elbow end of the row, though the number varies,
and occasionally two or three vestigial feathers may ap-
pear between the normal members and the missing
sockets. Single plumes are at times met with in front of
the row and are obviously representatives of a second row}
while in one farmer’s strain an almost complete second!
row of under-coverts occurs, alternating with the first,
and in front of this are five or six members of a third row.
One is forced to the conclusion that the ancestral ostrich
_ had the under surface of its wings provided with several
rows of under-coverts in the same manner as modern fly-
ing birds, and that the rare occurrences mentioned are in
the nature of survivals, the germinal factors responsible
for their appearance having been largely, though not yet
altogether, lost to the race.

The valuable wing quills or remiges ordinarily vary
from 33 to 39, having the same average, about 35.5, for
both the northern and southern birds. They constitute a
fluctuating series about the mode 36, though there is much
probability that each separate number in the series will be
found to represent a pure line. Assuming that not much
numerical variation occurs in the plumes of the ostrich
the farmer has never yet bred for quantity, quality has
been his only consideration. Recently however a cock
bird has been discovered among the government’s experi-
mental troops bearing 42 remiges, and it is submitted that
this high number represents an ancestral survival rather
than a reversion or mutation, and that the wing quills of
the African ostrich afford us various stages in degenera-

324 THE AMERICAN NATURALIST (Vou. LIT

wy

Fic. 1. Inder surface of wing of ostrich with plumes eee off. In by far
the majority of ostriches the surface is naked except for the sir :
under-coverts which is rarely complete; in the specimen represented six of the
coverts are missing from the elbow end of the row. In ome farmer’s strain an

he
q

bone
>
=
<4
<

ot

a third row. The third digit is almost buried in the flesh of the wing, but can
be seen projecting slightly towards its distal end. The claw which is present on
the first and second digits is not clearly shown

Fic. 2. Outer surface of wing of ostrich, the plumes having been clipped off
to show their arrangement in rows. The feathers in the uppermost row, the
a ng or ‘Feaniges, vary from 42 to 33 in different birds. The members of

ia ities, oki the eoid. row of coverts has often a easter missing towards
the free end of the row, though not in the wing egien The other rows of
coverts, third, fourth and fifth, may also show reduction. The marginal row of
the bastard wing may contain from two to seven feathers.

No. 627] GERM. PLASM OF THE OSTRICH 325

tion from the maximum 42 to the present minimum of 33.
As experiments have proved that the high number breeds
true, and as the other rows of commercial plumes vary in
correlation with the remiges, the discovery has a great
industrial bearing; for it now becomes possible to pro-
vide the farmer with a pure line of 42-plumed ostriches
in place of the degenerate 36-plumed birds with which he
farms to-day, and the entire feather crop will surpass the
present one by about 25 per cent.

The first row of upper-coverts varies in correlation
with the remiges (Fig. 2) but never shows any indepen-
dent reduction, while the second row has often a number
missing from its distal end, and is clearly undergoing
reduction here in contrast with the elbow end for the
under-coverts. Again, it is usually stated by writers that
the ostrich is destitute of an under-covering of down
feathers and filoplumes, yet in every northern and
southern bird examined, down in all stages of degenera-
tion occurs around the base of the larger plumes of the
wing and tail, and in rare cases spreads over a wider area,
leading to the conclusion that at one time the ostrich had
an under-covering of small feathers like flying birds
generally.

The third digit displays certain most unexpected evo-
lutionary stages. While in most cases it is altogether em-
bedded in the flesh of the wing, and can only be seen and
felt through the thin skin, yet occasionally its tip projects
quite freely, suggesting its former separation, like the
first digit which forms the ala spuria. Moreover, in some
birds odd feathers are to be found set along the finger,
altogether detached from any other series. These are
surely to be understood as survivals of a time when the
third finger was clawed, free and provided with its own
feathers, a primitive condition which is usually held to
be represented only in the oldest known fossil bird, |
Archeopterys. :

The legs and toes likewise exhibit degenerative phases.
The African ostrich is unique among living birds in hav-

326 THE AMERICAN NATURALIST [Vou. LILI

ing already lost its first, second and fifth toes, only the
third and fourth remaining. The outer, fourth toe is far
smaller than the inner third toe, and the condition of its

Re wae
tip a t
AA : I
re sy
aon xt
: ae
£ fe pia P
t em i wt i
T 3? ES
3 pa
i R Cy
I pa
»
L aa ion,
os 1g
g :
Se fi
— TaY
|S Aes
eee cA / C A
e ina 4] SA
Hr Sa
wo Mm” A mear EY
Sr Ge
~~ 5 y
yari TA
1 E A
. =~
ie TS { m.
JUI wK
Hi a YA zi Bee -mÁ
=A. j =
Ws = wh

s and foot of North African ostrich

Fic. 3. The sag paige toe is
greatly ede ap omparison with the

of scales a ntinuous. A second brea
joint of fre pa pias larger scales being divided into two or three smaller ones.

claw as well as its smaller size lead one to infer that it also
is on the road to disappearance (Fig. 3). In northern
birds the claw: of the fourth toe is frequently discernible,
though altog less, never reaching the ground;

No. 627] GERM, PLASM OF THE OSTRICH 327

but on only a few southern birds is it ever seen, and then
in a most vestigial state, barely showing beyond the skin.

What may doubtless be regarded as the first steps in
the degeneration of the big middle toe are also displayed.
Down the front of the tarsus extends a series of large,
nearly rectangular scales, continuous all the way from a
little below the ankle and passing over the upper surface
of the toe, though usually somewhat smaller where the toe
joins on to the tarsus. In a few ostriches a distinct break
occurs at the joint, several large scales being altogether
wanting (Fig. 3), and rarely birds are met with in which
a second break takes place over the middle joint of the
toe. One may hazard the suggestion that the interrup-
tion in the seutellation over the two joints has an adaptive
significance, allowing the parts to move more freely, but
we have also to face the fact that the single break occurs
in but a few while the double break is very rare. It is
presumably a new feature in course of introduction into
- the ostrich race, but not yet established for the members
asawhole. It involves however a reduction in the make-up
of the toe; it is a minus or retrogressive mutation, and
may well be the first hint of impending loss of what will
be the only toe when the small fourth has gone.

Although definite experimental data on all these reduc-
tion phenomena are not yet available everything points
to the fact that the variations breed true and are therefore
germinal in their nature; they are certainly not ordinary
fluctuating somatic variations. Proof is to hand that the
42-plumed cock has factorial representation for its high
number of plumes. Another similarly numbered hen is not
yet available, but in crosses with various 36-plumed hens
the average number of plumes of the progeny is 39.56
which is midway between the parents, while the mode is
40. Considering the heterozygous nature of the ostrich
where number of plumes is concerned a fluctuating series
of this kind is what would be expected. Only one farmer’ s
strain has the nearly complete second and third rows of
under-coverts, but they are found in all the progeny from

328 THE AMERICAN NATURALIST (Vou. LIII

the strain; all the members of a flock bred from the same -
stock have the second row of upper-coverts complete,
while in other flocks all the members have a number of
plumes absent from the row. Crossing of birds in which
the complete loss of the claw on the small toe has taken
place with others in which the claw still appears gives
results on strictly factorial lines, as also does the cross-
ing of birds with and without a loss of the scales. Ina
mixed assemblage of any species where only a small pro-
portion display a certain character it may be presumed
that the latter will be heterozygous with regard to the
particular character, seeing that the chances are much
against the mating of two individuals each having the
character. The heterozygous nature of the bird can be
demonstrated on mating with one in which the character
is absent, for if dominant it will appear in half the prog-
eny and be absent from the others. This proportion with
regard to the presence of the claw and the loss of the toe
scales has been found to hold in all the crosses. Out of a
total of 36 chicks hatched from breeding pairs where one
parent was clawed and not the other the numbers were
actually equal, namely, 18 chicks were clawed and 18 un-
clawed. Out of 11 chicks reared from a pair where one
parent showed no loss of scales on the big toe and the
other had a single break, 5 had no break and 6 showed the
break.

It may be accepted then that all the degenerative phases
represent factorial changes which have come about in the
germ plasm of the ostrich. Presumably the changes in-
volve a loss of factors; they are retrogressive or negative
mutations. Structures which would be expected to oceur
either fail to appear or are seen very rarely, and may
then be regarded as survivals, the factorial losses not
having yet taken place in the particular individuals.’
Thus, to take the case of the first row of under-coverts,
the principle of alternation demands that a complete row
of under-coverts should alternate with the row of remiges,
The full row actually occurs in a few individuals, and

No. 627] GERM PLASM OF THE OSTRICH 329

suffices to prove that this was the condition in the ances-
tral ostrich; more usually eight to ten are missing and
also fail to appear in the progeny. It is therefore reason-
able to assume that the germ factors originally involved
in the production of the eight to ten under-coverts have
disappeared from the majority of ostriches though they
are retained in a few. The fact that all the intermediate
numbers can yet be obtained shows the loss to have been
progressive. A similar line of argument can be applied
all through. Loss or degeneration is in progress in vari-
ous directions and differs in degree in different individ-
uals, and the losses are the outward expression of internal
changes in the germ plasm.

Where a loss of factors is taking place it could hardly
be expected that all the individuals of the race would be
affected at one and the same time. The process would be
more rapid in some than in others, some would incur the
loss at one time and some at another, and the results from
crossing would need to be reckoned with. Hence we can
understand the great diversity of stages represented in
the ostrich where large numbers are available for exam-
ination. It may be hard to comprehend how in the first
instance germinal changes can be brought about, but if
once effected, their repetition and continuance can reason-
ably be expected. Beginning with one or a few birds it is
manifest that as the loss in any direction continues more
and more individuals will become affected, until in the
end complete loss for the race will be achieved. So far as
the investigation of farmer’s troops has proceeded it
affords strong evidence for the view that only one orig-
inal 42-plumed bird now exists in South Africa, so that
under natural conditions the extinction of this high num-
_ber of remiges would be imminent. The loss of the claw
from the third finger is probably only recent. Some text-
books of zoology* assert that a third claw occurs, but it
has never been found on the hundreds of southern and
northern ostriches coming under my examination, al-
though specially looked for.

4 Parker and Haswell, Vol. II, p. 393.

330 THE AMERICAN NATURALIST [Vor. LIII

In most instances it would appear as if the loss of all
the many factors concerned in the production of a single
plume takes place simultaneously, as is the case with most
meristie structures; for usually the absences are complete
plumes. In some birds, however, two or three incom-
pletely formed or vestigial feathers occur between the
normal feathers of a row and the absent sockets, as if the
loss of the individual plume were taking place piecemeal.
This condition can be easily understood if we assume that
the constituent factors concerned do not all drop out
together, but follow some sort of succession. The factors
left at any time would then give rise to the part of the
feather for which alone they are responsible, and we
should get an imperfect or vestigial feather. In any
animal vestiges of a structure will continue to appear so
long as any of the factors concerned in the original struc-
ture remain. It is submitted that degeneration of any
complex structure never takes place in a gradual continu-
ous manner, as is usually supposed, but by successive
steps determined by the manner in which the factors drop
out; the appearance of continuity will however be con-
ferred if the steps are small enough. .

If a sufficient number of individual ostriches were
gathered together it could easily be made to appear as if
degeneration in any of the recognized directions were
taking place in a slow continuous manner, for all stages
between the extremes could be obtained. Proceeding by
such a method however would give an erroneous impres-
sion of what is actually happening. For although all
stages do occur they are in reality disconnected, and each
stage has been reached in an individual quite irrespective
of the others, and represents a separate and distinct
germinal loss; furthermore, in the same individual degen-
eration in any one direction proceeds quite independently
of the other directions in which the process is taking place.
It is not the wing as a whole which is undergoing degen-
eration, but the constituent parts of which it is made up,
each presumably represented by its own factors and be-

No. 627] GERM, PLASM OF THE OSTRICH 331

having with a large measure of independence. The losses
are continuous for the race but discontinuous for the indi-
vidual; and it is with the individual that heredity is con-
cerned and evolution with the race.

The degeneration phenomena presented by the ostrich
in connection with its wings and legs, as well as with its
plumage, would appear to provide us with an example of
the application of mutative and Mendelian principles to
such evolutionary facts as confront the comparative anat-
omist and paleontologist. So far as concerns the indi-
vidual bird the retrogressive changes are shown to occur |
as separate mutations and to follow definite factorial
lines, while as concerns the evolution of the race they pro-
ceed in a continuous determinate manner. In all prob-
ability they take place wholly irrespective of any adaptive
significance or consideration for the welfare of the bird,
and are intrinsic in their nature and uninfluenced by ex-
ternal conditions. Natural selection has probably played
no part in connection with the losses, for the greater
changes have already affected the race uniformly and the
smaller ones which still vary in degree in different indi-
viduals will probably affect the whole in the end. Should
the loss of plumage continue to a much further degree
and marked degenerative changes be set up in the big
middle toe natural selection may then be expected to bring
about extinction.

The chief point desired to establish at present is that
as regards the number of its wing plumes and in certain
other features the ostrich affords strong support for the
view that its hereditary determiners or factors are chang-
ing regularly and frequently; they are not fixed and con-
stant as are the factors for the structural details of the
plumes; one series is in a state of change, the other is non-
changing. The great variety and degree of the degenera-
tive stages in the ostrich of to-day admits of no question,
and that they are the expression of so many germinal
differences may be accepted, seeing that they breed true;
that they have been effected simultaneously as we find

332 THE AMERICAN NATURALIST (Vor LIL

them is inconceivable, and we are justified in concluding
that in the past the germ plasm has changed frequently
and presumably over a long period. Moreover, we can
hardly admit that the various degenerative phases will
remain as they are at present, but that further losses in
the same direction will follow, that is, the germ plasm will
continue to undergo retrogressive changes of a like
character to those already initiated. We may have an
appearance of continuous change, but when analyzed it
will be found to proceed by means of separate factorial
steps. It is conceivable that a continuance of the kind of
factorial losses now in progress will result ultimately in
the complete disappearance of the wings and legs of the
ostrich, allowing that the bird could survive the inter-
mediate stages, a postulate which it must be conceded is
of no mean order. May we not suppose that the limbless
condition of snakes and some lizards, amphibians and
fishes has come about by the successive losses of germinal
factors in a similar manner to that which is here shown to
be taking place in the ostrich?

The bearing of the germinal changes iavobved in the
degenerative processes upon the thesis of Section IV, may
be noted. It is in the highest degree improbable that
determinate losses of such a widely embracive nature are
taking place in response to any environmental stress
acting upon the germ cells; rather they may be regarded
as the result of some wholly internal physiological cause.
If due to environmental stress one could-reasonably ex-
pect that in any individual the losses would be taking
place in all directions simultaneously, and would have
reached about the same degree in all. But among the
various rows of feathers, as well as in other parts, the
greatest independence in reduction is met with, as if the
factors for each were subject to a separate rather than a
common influence.
VI

“‘ Mixing of germ plasms in fertilization alters hered-
itary determiners mutually and hence is, in and of itself,

No. 627] GERM, PLASM OF THE OSTRICH 333

a cause of genetic variations . . . a purely external agent,
the continued selection of personal somatic qualities, will
alter the germ plasm.’’

It seems to be generally allowed that at any period the
majority of forms of life are static so far as germinal
alterations are concerned, while some are undergoing
progressive changes and others retrogressive changes.
During the present period the representatives of the
widely distributed Ratite are unquestionably undergoing
marked changes and have been for ages past. The
changes are in a negative or retrogressive direction, and
express themselves in somatic degeneration, particularly
with regard to the wing and shoulder girdle. The living
Apteryx is a well-known instance of wing degeneration, as
also the recently extinct moas, in which no hint of a wing
has been found and a trace of the glenoid cavity only in
one species. From the data already submitted we are
able to learn something as to the manner in which the
degenerative processes are proceeding in the wing and leg
of the ostrich, and presumably the same method holds for
the Ratite generally. Factors are evidently in process of
dropping out, in regular succession, along definite pre-
scribed lines, the degree varying much when the entire
race is taken into account.

All Mendelian writers seem to concede that the fac-
torial changes, plus or minus, are not autonomous on the
part of the factors, but are ‘‘a result either of wholly
internal physiological causes, or of very extraordinary
environmental stresses acting directly upon the germ
cells.” Though the results of Morgan and his associates
indicate that it may yet be possible to understand the
manner in which the factors undergo their changes it will
always be competent for us to enquire as to the cause or
causes inducing the changes. To be complete our analysis
of variability will need to get beyond the factors to the
force or forces acting upon them. In the previous sec-
tions good reason has been adduced for supposing that
the losses going on in the germ plasm of the ostrich are

334 _ THE AMERICAN NATURALIST [Vou. LIII

due to some wholly intrinsic cause, and seeing that it
affects all the members of the race and has been operative
for a long period we may conclude that it is transmissible
and acts continuously from generation to generation.
The many stages represented also give some justification
for supposing that whatever the cause of the factorial
changes may be it varies in intensity in different mem-
bers of the race, being less active in individuals where the
loss of plumes is small as compared with others in which
the loss is greater. For example, the causative agent
bringing about the loss of the plume factors must be less
in intensity or less active in 42-plumed ostriches than in
33-plumed birds. We may with good reason expect that
the selection for breeding of the high numbered birds
will arrest the rate of degeneration of the race in this
` particular feature, while on the other hand the selection
of the low-numbered birds will tend to accelerate the rate
at which the factorial losses are taking place. Where
therefore the germ plasm of a race is in a continuously
changing phase, as in the ostrich, we can hope to retard or
accelerate the changes by selecting individuals differing
in the degree to which they are under the influence of the
causative agent. It is submitted that in this sense we can
say that ‘‘a purely external agent, the continued selec-
tion of personal somatic qualities, will alter the germ
plasm.’’

‘We can not hope that the continued selection of 42-
plumed birds will in the end give to the farmer ostriches
with a still higher number of remiges, as the factors for
the plumes beyond these have in all probability disap-
peared from the race, and there is no evidence that the
cause of the factorial changes is effective in a plus but
only in a minus direction. On the other hand the con-

tinued selection of 33-plumed birds may reasonably be ex-
: ed to accelerate the loss of the remiges, by leading to
a more -rapid loss of the factors. Owing to the present
degenerative forces at work in the ostrich we can by
selection hope to modify the germ plasm in a minus direc-

No. 627] GERM PLASM OF THE OSTRICH 335

tion, though not in a plus direction, beyond the present
limits of the race. It will of course be readily appre-
ciated that this possibility differs altogether from that
due to the ordinary selection which may go on in a race of
organisms where the germ plasm is static, but where all
grades of pure lines may be extracted between extreme
limits. Where the germ plasm for a race is static, as
demonstrated by Jennings in his work on Paramecium,
we can readily understand that no further change is pos-
sible by selection within a pure line, as nothing inducing
factorial changes is present. If where germinal changes ~
are taking place it is not permissible to think of the
factors as changing autonomously we have to assume that
some causative agent is present, and may vary in degree
in different members and thereby form a basis for selec-
tive action.

The same considerations can be applied to the state-
ment: ‘‘mixing of germ plasms in fertilization alters
hereditary determiners mutually and hence is, in and of
itself, a cause of genetic variations.” When, for ex-
ample, two germ plasms, in each of which the causative
agent producing loss of factors is at its maximum, become
mixed in fertilization it is reasonable to expect that the
agent will be intensified and the hereditary determiners
will be altered mutually, and some of them drop out. The
mixing will be, in and of itself, a cause of genetic varia-
tion, which will be expressed by a further loss of remiges.

Though the idea of a causative agent inducing changes
in the germ plasm, and varying in degree and also trans-
missible, is altogether hypothetical yet it is stimulating
to further experimental effort. Of the hundreds of
ostriches examined not one has been found with less than

` 33 remiges, hence this number must be regarded as the
present minimum of the race. There is every reason to
expect that a pure line having this: number only can be
built up. If by breeding these together a further reduc-
tion of plumes should take place we should then be fully
justified in assuming that the factors concerned with the

336 THE AMERICAN NATURALIST [Vou. LII

lost members had dropped out from the germ plasm,
especially if later breeding failed to restore them; selec-
tion would have induced a definite change in the germ
plasm. Also if a pure line with 42 remiges were estab-
lished and no further increase oceurred we should be
warranted in concluding that the factors for the plumes
beyond this number had already disappeared from the
race and could not be restored; the causative agent could
not act in a plus direction. It is unfortunate for such in-
vestigations that the ostrich is such a slow breeder. Ex-
periments are however being conducted to determine if
further reduction in the 33-plumed birds can be induced,
while the building up of a pure 42-plumed race is also in
progress, the latter having an important industrial bearing.

In many respects the degeneration phenomena in the
ostrich appear to be best understood on the conception of
autonomous changes and variations in potency of the
germ factors. In the case of the dropping out of plumes
during the chick stage the reduction in potency has pro-
ceeded so far as to result in entire loss of effectiveness
only from the chick stage onwards, while complete loss
of factors from the germ plasm may be regarded as the
final loss of potency. May not a variation of potency of
factors be at the root of many of the so-called fluctuating
variations? The explanation seeks for the loss of factors
among the autonomous changes in the factors themselves,
while the idea of a ‘‘causative agent’’ throws the respon-
sibility for the changes upon some influence external to
the factors.

Since the above was written certain results have been
obtained which strongly support the idea that it may be
possible to induce retrogressive changes in the ostrich.
As stated, a loss of scales over the large middle toe has’
already taken place in a small proportion. Out of twenty
southern birds of mixed breeding one showed a single
break while out of twenty mixed northern birds a single
break occurred in three cases and a double break in two.
The results given below are derived from the mating of a

No. 627] GERM PLASM OF THE OSTRICH 337

northern cock without any break and a southern hen in
which the break occurs. Of the four offspring reared
three are without the break while it occurs in hen No. 179.
From the mating of brother and sister four F, chicks
were hatched, two of which have a double break in the
scutellation, one shows a single break and one has no
break. Thus the proportional loss of scales has greatly
increased in the F, generation.

SCUTELLATION IN F, Cuicks Comparep Wirth Parents
AND GRANDPARENTS

No Break Break
NN. Ae OSE INO. Dire eto acs te eats X —-
SoA bon, No DaD cise a — x
F. Crosses.
Cross-bred cock, No. 182............. X ==
Cross-bred hen, No. 179 .........-...-. — x
F; Chicks.
ONG R as py VEE TE PETE N — xx
No: ied oe ger ee OE TE hee — x
NO. Bo sve ewes ees ge E oes s8 hs —_ xx
E ae Stet SiG CEES. iene. T x —

From what has been adduced already we may with good
reason admit that an inherent tendency exists in the
ostrich towards the loss of certain parts of the fore and
hind limbs, and the above result may be regarded as
highly suggestive that by inbreeding the inherent tend-
ency towards the loss of scales can be accentuated along
definite lines. The accumulation of fuller data must be
awaited before the suggestion can be regarded as more
than tentative.

ADAPTATION AND THE PROBLEM OF ‘‘ORGANIC
PURPOSEFULNESS.’’ II

DR. FRANCIS B. SUMNER

Scripps INSTITUTION FOR BIOLOGICAL RESEARCH, LA JOLLA, CALIF.

IV. Tue PRINCIPLE or TRIAL AND ERROR In RELATION TO
REGULATIVE PHENOMENA!‘

Driesch and some other vitalists draw their most ef-
fective ammunition from the phenomena of experimental
embryology and regeneration. How is it that a frag-
ment of a developing organism—any fragment, within
certain limits—can produce the whole? How it is that
various perversions of the normal course of development
do not prevent the attainment of the normal end? How
is it that certain adult organs, e. g., the lens of the eye of
a triton, when removed by a highly ‘‘unnatural’’ opera-
tion, is nevertheless restored, and restored by a process
quite different from that in which it is normally pro-
duced in embryonic development?

At the outset we must make two admissions: (1) that
these processes can not be the result of a mechanism spe-
cifically adapted in advance to meet these particular exi-
gencies, and (2) that they can not be satisfactorily
explained by assuming any preformation of the parts
which are restored. The former supposition is to be re-

14 The ‘‘trial and error’’ principle na of late years come into the fore-
ground of biological discussion, largely through the writings of Jennings.
It was, so far as I know, first clearly proposed PIRR not so named) by
Spencer (Principles of Fisddose, Vol. I, pp. 544-545) to account for the
origin of adaptive responses to stimuli, and was later developed by Bain.
There are important points of agreement between the views of these writers
and some of those set forth independently by Roux in his classic essay,
‘*Der Kampf der Theile im Organismus’’ (1881). More recently, Baldwin
(Mental Development, 1898, Chapter VII; Development and Evolution, 1902,
pp. 108-115) has further elaborated the same fundamental idea as that of
Spencer and Bain in his theory of ‘‘ functional selection.’’ Various animal
es. (e.g:, Lloyd Morgan and Thorndike) have also laid stress on
this principle.

338

No. 627] ADAPTATION 339

jected on account of the unusual and artificial character
of the operations, which could never have been provided
for by natural selection, nor, so far as we can see, by any
other recognized principle of evolution. The latter sup-
position is sufficiently disposed of by Driesch’s analysis
(section III) and need not be considered here.

Driesch admits that a physico-chemical machine ‘‘might
very well be the motive force of organogenesis in gen-
eral, if only normal, that is to say, if only undisturbed
development existed, and if a taking away of parts of
our systems led to fragmental development” (II, 139).
If, therefore, we can explain these critical cases without
invoking any principles beyond those believed to be oper-
ative in normal life-history, we have disposed of this line
of argument.

In an earlier section of this paper I took the ground
that an adaptive or ‘‘purposive’’ response by the organ-
ism, if not guided by past individual or racial experience,
must be the result of experimentation. I avoided inten-
tionally at the time any consideration of those cases of
regeneration and form regulation in which the emer-
gency was totally new, and therefore foreign to the expe-
rience of the organism or its ancestors. Here a specially
evolved mechanism could hardly be invoked. I sug-
gested, however, that the principle of ‘‘trial and error”
could be applied to these cases: This suggestion was, of
course, not new. Such an extension of this conception
had already been made by Jennings,’® though it is rather
surprising to note that he has given it little further con-
sideration in his recent discussions of vitalism. For, to
my mind, an explanation involving this principle, seems
the only alternative at present to a vitalistic one, or,
better stated, it seems to me the only alternative to an
abandonment of the search for a scientific explanation.

According to the trial and error principle, as applied
to the movements of a lower animal, ‘‘behavior that re-
sults in interference with the normal metabolic processes

15 ‘ Behavior of the Lower Organisms’’ (1906), Chapter XXI.

340 THE AMERICAN NATURALIST (Von. LIII

is changed, the movement being reversed, while behavior .
that does not result in interference or that favors the
metabolic processes is continued.’"® The primary
“avoiding reaction,’’ in the presence of an unfavorable
stimulus, is, of course, comparable with a simple reflex.
Its ordinary effect is to remove the organism from the
noxious influence. When progressive movements are re-
sumed, they occur at random, so far as their direction is
concerned, and they may or may not take the organism
into favorable surroundings. If they chance so to do,
they are continued indefinitely. If not, the reversal of
movement.occurs as before. Thus while, to the uncrit-
ical observer, the organism seems to ‘‘seek out’’ the opti-
mum environment, it really reaches this through a series
of accidents. This is as true of a cat, releasing itself
from an experimental trap, as it is of a paramoecium
escaping from a harmful to an optimum water tempera-
ture. In the case of the cat we may be tolerably sure
that the animal experiences a feeling of discomfort until
the means of escape is discovered, and we find it conve-
nient, if not inevitable, to say that her restless move-
‘ments are the result of this feeling. In the case of the
infusorian, we are much less sure of the conscious ele-
ment, though its introduction is permissible as an act
of philosophic faith. In theory, most scientists are prob-
ably psychophysical parallelists, but in practise it seems
necessary at times to use the language of interactionism.
In discussing the voluntary movements of a higher ani-
mal, any other course would seem pedantic. But in dis-
cussing the simple behavior of a lower organism, such
language is commonly branded as ‘‘anthropomorphic.”’
Nevertheless, I believe that its employment even here is
sometimes useful in forcing us to keep in view the essen-
tial unity of animal life. No protest is raised by the
physiologist when thoroughly protozoomorphic language
is applied to a vertebrate. Why then should ‘‘anthropo-
morphic’’ terminology so shock us in describing the be-

16 Jennings, op. cit., p. 39.

No. 627] ADAPTATION 341

havior of a Paramecium? Each is the extension of an
article of philosophic faith far beyond the realm of ex-
perience. But this is no essential part of our present
argument. Let us consider whether the trial and error
principle may not be applicable to other phenomena than
the bodily movements of animals.
Jennings asks:
s it possible that interference with the physiological processes may

and the like,—and that one of these activities is selected, as in behavior,
through the fact that it relieves the interference that caused the change?

. It is evident, then, that the organism has presented to it, by the
EENT just sketched, unlimited possibilities for the selection of dif-
ferent chemical processes. y is a great mass of the most varied
chemicals, and in this mass SS of chemical processes, in every
direction, —all those indeed that are possible—are occurring at all
times. There is then no diffculty as to the sufficiency of the material
presented for selection, if some means may be found for selecting it
(op. cit., p. 346).

Looking for evidence that such a process of selection
does actually occur in physiological regulation, Jennings
cites the experiments of Pawlow, in which the latter
habituated dogs to various kinds of foods and noted the
effects upon the digestive juices. In these experiments
the adaptive changes in the activities of the digestive glands, fitting the
digestive juices to the food taken, do not occur at once and completely
under a given diet, but are brought about gradually. . . . This slow
adaptation is, of course, what should be expected if the process occurs
-in anything like the manner we have sketched (p. 347).

Jennings concedes:

It is perhaps more difficult to apply the method of regulation above
set forth to processes of growth and regeneration. Yet there is no
logical difficulty in its way. The only question would be that of fact,
whether the varied growth processes necessarily do, primitively, occur
under conditions that interfere with the physiological processes. When
a wound is made or an organ removed, is the growth process which fol-
lows always of a certain stereotyped character, or are there variations?

It is well known, of course, that the latter is the case. . . . Removal
of an organ is known to produce great disturbances of most of the
processes in the organism and among others in the pro of gro

. . . Some of these relieve the disturbance; the variation then ceases
and these processes are continued (p.

342 THE AMERICAN NATURALIST [Vou. LIII

A line of argument which has points of similarity to
the foregoing has been independently developed by
Holmes.‘* He believes:

The harmonious functioning of an organism is mainly secured by a
system of automatically acting checks which we may conceive to act
in manner more or less remotely analogous to the governor fe a steam-
engine or the forces which regulate thé motions of the plane
these cases deviation from the normal is the cause which iene
sets up activities by which the normal is regained.

So, too,
the self-regulation of organisms may . . . be in a measure understood
if we assume that their parts stand in a SS of mutual dependence
such that the undue growth or functioning of any part is held in check
by the reactions thus brought about by other, and especially the con-
tiguous structures. If we suppose that the various cells constituting
the body have each a different kind of metabolism, and that the products
of each cell are in some way utilized by the neighboring cells, so that
each derives an advantage from the particular association in which it
occurs, we may understand, in a measure, how this checking may be
brought about.

And here an analogy is pointed out with the relations
which obtain in ‘‘symbiotic’’? communities, such as those
composed of animal cells and certain unicellular alge.

The conception here developed is in some respects an
extension of Roux’s intra-selection hypothesis, though
Holmes rejects the notion of a ‘‘struggle of the parts.’’
This conception, which derives strong support from re-
cent discoveries respecting ‘‘hormones,’’ gives a certain
measure of concreteness to that rather vague expression,
“the organism as a whole.’ For, despite the many |
known instances of local autonomy, we can not doubt
that the organism does in a high degree act as a whole.
But this ‘‘ wholeness’’ may not be an irresolvable fact, as
has sometimes been assumed. It may be possible to con-
ceive it in terms of chemical and structural integration. "°

This hypothesis, as applied to form regulation, would

17 Archiv fiir Entwicklungsmechanik, 1904

17a To me, such a viewpoint seems quite teeoneilable with the ‘‘ organis-
mal’’ conception of Ritter, though Professor Ritter himself (The Unity of the

Organism, Vol. I, p. 183) has gone to considerable pains to show the fallacy
of Holmes’s ‘position

No. 627] ADAPTATION 343

seem to be closely related to that of Jennings, and in-
deed Jennings himself views it in this light. It is diffi-
cult to gather, however, to what extent Holmes has in
mind the principle of ‘‘trial and error.” His compari-
son of regeneration with functional hypertrophy does
not seem compatible with this principle. ‘‘Remove one
of a pair of organs,’’ he says, ‘‘and its fellow increases
in size. Remove a part of one of these organs and the
remaining portion grows, forms new tissue, and regen-
erates the missing part.’ Furthermore, he believes that
these phenomena may be analogous with some of those
described under the name of ‘‘chemical equilibrium.’’

The decomposition of compounds in solution proceeds until there is a
definite relation established between the amounts of the old compounds
and the new. If the chemical equilibrium thus established is disturbed
by the removal of one of these compounds more of that compound will
be produced; and the more rapidly the compound is removed, the more
rapidly it is formed.

Such an ‘‘automatic’’ restoration of equilibrium as
this might seem to be a radically different thing from
trial and error. The process by which it is attained
would appear to be direct and unhesitating. Holmes
says that the solar system, no less than the organism, is
a‘‘self-regulating mechanism.’’ Now, in the former, the
balance of its opposing forces is effected ‘‘automati-
cally” in the sense that any deviation in the movement
of one of the parts would result inevitably in a compen-
sating deviation in the others. Is the restoration of an
organism to ifs norm of this direct and automatic type?
Are such processes as tend to compensate a disturbance
in the normal functioning of an organism the direct and
exclusive result of the disturbance itself, or does this dis-
turbance evoke a variety of responses of which the suit-
able response may finally happen to be one? The first
of these alternatives may be admitted as probable in the
ease of such disturbing factors as have been frequently
experienced in the past. But how does it happen that
certain cells of the iris of a newt become stimulated to
division by the removal of the lens? And why should

e
344 THE AMERICAN NATURALIST (Vou. LIII

their metabolism become so affected that they give rise
to lens tissue, instead of to iris tissue? Can we believe
that the iris cells proceeded unfalteringly to this end as
a result of the operation?

The discussion after all hinges upon the word ‘‘unfal-
teringly,’’ and this term has been applied to processes
which are beyond the possibility of direct observation.
If we grant that a disturbance of growth equilibrium was
what led to the reparative processes, and that equilib-
rium was in the end restored, it does not seem difficult
to admit that each minutest step in the direction of re-
storing this equilibrium was selected from a medley of
random reactions. Indeed, Holmes suggests that
cells which develop in the direction of the missing part receive those
advantages which the symbiotic relation afforded the cells whose place
they take. Differentiation in any other direction deprives them of these
advantages and subjects them to other unfavorable conditions.

Nor need it be assumed that these responses are wholly
random. Although it is incredible that each type of
possible injury has been provided for in advance by a
specific mechanism, it seems more than possible that cer-
tain reactions have been acquired which are of service
in any emergency—a sort of ‘‘first aid to the injured,”
as we might say. After these preliminary steps of a
general character—which are, as a matter of fact, the
common precursors of regeneration'’—the more special
processes may be supposed to proceed in a tentative
fashion.

All that is meant by ‘‘growth equilibrium,’’ in this dis-
cussion, is such a normal state of metabolic balance that
the growth of each part is checked through its organic
relations with the rest. Attainment of this goal would
bring the organism into a condition of ‘‘no stimulation,”
like that of the protozoan which has escaped from an un-
favorable environment.

Since we commonly are able to observe only the final
outcome of such a process, and overlook the minute steps

18 These steps are frequently retrogressive ones and include the loss of
specialized structures.

No. 627] ADAPTATION 345

by which it comes to pass, we are wont to believe that the
reparative activities move directly toward the end which
we observe to be ultimately attained. Thus Driesch tells
us that

the process of restitution, perfect the very first time it occurs, . . . is
the classical instance against this new sort of contingency. .. . Here
we see with our own eyes that the organism can do more than simply
perpetuate variations which have occurred at random.

What we see with our own eyes, as I have already said,
is only a series of visible stages in the process of resti-
tution. We do not see the inmost morphogenetic proc-
esses, physical and chemical, by which this end is at-
tained.

Perhaps it may seem that the foregoing explanation
merely resorts to the familiar expedient of throwing our
difficulties back into an invisible realm where they are
safely beyond the reach of scientific investigation. I
would say first of all that even this type of explanation,
which at least speaks in the language of known facts, is
preferable to one which frankly abandons scientific prin-
ciples altogether. And secondly, I would point out once
more the possibility that this hypothesis is one which
may in reality be put to experimental test. For any in-
dication of a profiting ‘by ‘‘experience,’’ i. e., of a short-
ening of the time required to effect a given regulative
response, would harmonize well with the hypothesis that
the response was at first effected through tentative steps.
Indeed, such evidence, even now, is not wholly lacking.

It may be well to remind ourselves at this point that
the perfect regeneration of missing parts, or the com-
plete reconstruction of a mutilated embryo is after all an
exceptional phenomenon. Many animals almost entirely
lack the power of regeneration, while most injured eggs
either die or give rise to abnormal embryos. These facts
harmonize best with the view that regenerative processes
are causally produced in the same sense as inorganic
phenomena, and that they are not determined, in any di-
rect way, by needs or ends to be realized. The forma-

346 THE AMERICAN NATURALIST [Vot LOI

tion of misplaced, supernumerary and other useless
structures, and the occurrence of anaphylaxis, instead of
immunization, certainly do not argue for the existence
of a ‘‘primary teleology’’ in nature, though, of course,
they do not wholly refute it.

On the other hand, the occurrence of these non-adaptive
responses to growth stimuli is no more inconsistent with
an intra-selection hypothesis, such as that here advo-
cated, than is the occurrence of multitudes of non-
adaptive structures or colors in nature inconsistent with
the theory of natural selection. There must be rigid
limitations to the operation of both processes. The task
which I have undertaken here is not to explain structures
and function in general, but the more modest one of try-
ing to explain why certain among these are directed to-
ward the conservation of the individual or the species.
If various other vital phenomena are found to be non-
adaptive, our difficulties ought not to be increased.

There are cases, it is true, in which some simple phys-
ical factor, such as gravity, or the plane of section, may
determine whether the actual missing part is restored or
a misplaced organ is the result. It certainly seems arbi-
trary to offer fundamentally different explanations in
the two cases. Now, I have nowhere made the conten-
tion that the processes involved in regeneration are
wholly random, in the sense of being unrelated to one
another and to the past history of the individual. In
normal development the processes are doubtless so con-
catenated that growth and differentiation proceed in a
direct way with little or no ‘‘lost motion.” And every
detached portion of such an organism must receive its
share of this established developmental machinery. The
tendency to reconstruct the whole, to attain the normal
specific form, is therefore opposed by another set of
tendencies, urging it to develop as if it were still part of
the undivided organism. As is well known, the outcome
of this conflict of forces varies, depending upon the spe-
cies of animal and the time of operation. We may have

No. 627] ADAPTATION 347

either total or fractional development as a result. It
does not seem unlikely, therefore, that in every case of
regeneration the control of the ‘‘organism as a whole’’ is
opposed, more or less successfully, by the specific growth
tendencies of the various cells and tissues from which
restitution proceeds. These might, in consequence, bring
about the ‘‘autonomous’’ production of a wholly mis-.
placed part.1® Thus the phenomena of ‘‘heteromorpho-
sis’? should seem to offer no insuperable obstacle to the
views herein set forth.

Applied to the ordinary phenomena of regeneration,
say to the restoration of an amputated limb, or even the
lens of an eye, this hypothesis of achievement through
experimentation would seem to make no impossible de-
mands upon our imagination. We need only suppose
that the absence of the missing part serves as a stimulus
to varied and undirected metabolic activities, that such
of these as serve to restore the normal condition tend to
be continued and that growth equilibrium (absence of
stimulus to growth) is not normally attained until the
missing part is restored. The case would seem to be not
very different from that of an animal finding its way out
of an unfavorable environment. In both instances we
may suppose the organism to be in a condition of ‘‘un-
rest’’ until the end is achieved. This condition may or
may not be conceived in psychical terms. If so con-
ceived, the notion would be philosophically legitimate,
though scientifically unnecessary.”

When, however, we consider Driesch’s crucial case of
the development of an entire organism from an em-
bryonic fragment, the matter is admittedly far less con-
ceivable. For this fragment has retained nearly or
quite the same potentialities as the entire egg or embryo,
in that its career of multiplication and growth is brought

19 This explanation of heteromorphosis is, I think, quite in harmony with
that offered by. Holmes (op. cit., pp. 302-303).

20 Cf, Baldwin’s statement (‘‘Mental Development,’’ p. 177): ‘‘the life-
history of organisms involves from "e start the presence of the organic
analogue of the hedonic consciousness.

348 THE AMERICAN NATURALIST (Vou. LILII

to a close only through the attainment of the form which
is typical for the species in question. Why should this
ultimate condition of equilibrium be the same whether
we start from an isolated blastomere, an irregular frag-
ment of a blastula or anormal egg? Does it not seem as
if the only constant feature in this case were the end
itself? In considering the behavior of a protozoan, the
stimuli may vary and the method of escape may vary,
but the organism itself is the same. The ‘‘equi-finality’’
of the result—to use an expression of Driesch’s—may be
attributed to this fact that we are dealing with the same
physico-chemical system, and one of the self-regulating
type. But what of our various embryonic fragments?
Are they not obviously different physico-chemical sys-
tems?

Now, after all, the difference between this case and
that of a regenerating limb or lens appears to me to be
only one of degree. The distinctions relate (1) to the
stage in development at which the injury is inflicted, and
(2) to the proportional part of the organism which is left
to reconstruct the remainder.

1. As regards the first point, we must suppose that at
each stage of ontogeny such a state of physiological bal-
ance is normally maintained as is appropriate to that
particular stage. That the multiplication and differen-
tiation of certain cells is profoundly influenced by the
presence or absence of other cells is one of the assured
results of experimental embryology. One need only cite
the difference between the development undergone by an
amphibian blastomere which is totally detached at the
two-celled stage, and that of the blastomere whose part-
ner has been injured by a needle-prick and left in po-
sition. ia

Thus we have as much right to assume for the blastula
as for the adult animal that any disturbance of metabolic
balance will be followed by varied responses, some of
which will tend to restore the balance normal to that
period. The fact that these responses are known to
differ radically, following the same type of operation,

No. 627] ADAPTATION 349

and that the result is often a very imperfect reconstruc-
tion of the whole, lends support to the view that the cells
of the injured embryo ‘‘feel their way’’—so to speak—
back into a condition of mutual equilibrium. In some
cases this equilibrium appears to be of a simple physical
sort, as for instance, that which is brought about by the
folding together of the edges of a blastula fragment so
as to reconstruct the spherical form. But in most cases
the factors are doubtless vastly more complex.

Once the reconstruction of the normal embryonic form
is attained, the difficulties in understanding the further
stages of ontogeny are no greater than we meet with in
the case of an uninjured embryo—that is, unless we are
encumbered by a preformation theory of development.

2. As regards the second point above raised, there is
theoretically no greater difficulty in understanding how
one tenth of an organism may restore the remaining nine
tenths than in understanding how the nine tenths may
restore the one tenth. As a matter of fact, in dealing
with certain organisms, the size or shape of the piece, or
the region of the body from which it is taken count for
little in the outcome. But they do count for something,
and that something is significant. It has been found in
some cases, for example, that there are lower limits to
the side of the pieces which may carry out development
or regeneration. And in other cases, the position of the
plane of section may determine whether a useful struc-
ture is formed or one which is wholly useless.

But whether or not the size or shape of the fragment
count for anything in the reparation of a given organism,
we find that the species from which it is taken counts for
everything. There must, therefore, be something that
is common to all detached portions of an organism which
are capable of reconstructing the same whole. The por-
tion in question may be an asexual spore or a fertilized
egg, or it may be an isolated blastomere or other arti-
ficially detached fragment of either an embryo or adult
organism. What is this greatest common divisor? Is
it a unit of structure or is it a chemical substance?

350 THE AMERICAN NATURALIST [Vou. LII

There would seem to be no third possibility, as long as
we keep within the bounds of scientific explanation. But
a unit of structure may none the less be itself a chemical
individual. Modern speculative physics refers all quali-
tative differences in the last resort to differences of struc-
ture, even in the case of the elements. And it has been
suggested that the various specific protoplasms, which
are responsible for the slightly different metabolic prod-
ucts of different species, owe their differences to stereo-
isomers, i. e., substances which agree quantitatively in
their composition, but whose enormously complex mole-
cules differ as the result of some slight transposition of
atoms or radicals.?!

To the majority of present-day geneticists there is
doubtless a ready answer to the question: what is this
something that is common to all detached portions of an
organism which are capable of reconstructing the same
whole? It is likely that to most of them a completely
satisfactory answer would be: the cell nucleus. Thus
Jennings," in discussing specifically certain of the ques-
tions raised by Driesch, assures us that ‘‘the recent
study of genetics has shown that this [the chromosomal]
apparatus is the system on which the peculiarities of
development mainly depend. This system is not equi-
potential; the fate of its parts is not a function of their
position; it has a complex structure with a correspond-
ing complexity of action; altering any of its parts alters
correspondingly the action of the system; irregular re-
moval or disarrangement of the parts destroys the
action.”

Whether or not this aggregate chromatin matter of the
nucleus constitutes the minimum divisible of the organ-
ism, as recent students of heredity are disposed to believe,
is still quite undecided. For protozoa we are definitely
able to state that this is not true. Experiments in regen-

21 Reichert, Science, November 6, 1914. This article contains much inter-
esting evidence for the enian distinctness of genera and species, and

even of individual organism
21a Philosophical Review, Nor, 1918, p. 586.

No. 627] ADAPTATION 351

eration show that there must be smaller bodies within the
nucleus, each containing the potentialities of the entire
organism. Ritter?!» has recently insisted that the con-
cept of heredity must be applied unreservedly to these
one-celled organisms, many of which are quite complex
in structure and undergo a true ontogeny. Indeed, the
experimental studies of Jennings and his students have
demonstrated the transmission of individual peculiar-
ities, both of structure and function. As for the metazoa,
despite the considerable evidence for chromosomal —
viduality’’ and for the localization of genetic ‘‘factors,’
it seems to be entirely premature for us to assume the
existence of a mosaic of parts, rigidly predetermined and
incapable of making good a loss. One should recall what
happened to an earlier ‘‘mosaic theory’’ of development.
To go to the other extreme, it might be supposed that
for each form of organism there was at least one sub-
stance, or molecular structure, which was typical for it,
and which determined its specific physical and chemical
characteristics. The other constituents of the adult body
would be modifications of this typical substanee, which
had lost certain of its original components or acquired
new ones. This specific protoplasm would have some
points in common with the ‘‘germ plasm’’ of Weismann.
It might be credited with the power of indefinite growth
and self-division, so long as these were not checked by
counterbalancing forces. When completely checked, a
growth equilibrium would be established which would
represent the normal form of the species in question.
The rather vague and indefinite point of view here sug-
gested would avoid, however, the tangle of unverified
assumptions that are involved in the hypothesis of a
‘‘yerm-plasm,’’ conceived as an aggregation either of
Weismannian ‘determinants’? or twentieth-century
‘“venes.’? The admitted possibility that certain mate-
rial particles of the nucleus are functionally related to
separately heritable adult characters does not constitute
21b The Unity of the Organism, Chapt. XII, XIII.

352 THE AMERICAN NATURALIST [Vou. LIII

a proof that the entire organism develops through the
combined activities of such particles. Moreover, even if
such a complete germinal representation of adult char-
acters were shown to exist, only a part—and a minor
part—of our difficulties would be solved. We should still
have to explain how the elementary parts of the body
came to arrange themselves in proper spatial order and
in proper chronological sequence during development.
Blocks do not build themselves into houses. Driesch
points out that historically vitalism and epigenesis have
always been closely related, while the mechanistic school
has commonly adopted some form of preformationism.
Such a connection is far from being logically necessary,
however. To me it would seem that preformation lent
itself most readily to vitalism—to the notion of a builder
who put the blocks together. In our particulate theories
of organic differentiation, we commonly leave out of ac-
count the spatial and chronological relationships of the
parts, or rather we take them for granted. We assume
that somehow our ‘‘organismules’’ will find their way to
their proper places at the proper moments, just as in a
laboratory experiment the experimenter himself sees to
it that everything is at each moment just where it be-
longs. 5

Let us return to an illustrative case, already consid-
ered, and ask why no one has ever seriously proposed a
. preformation theory of the earth’s origin. Most mod-
erns (M. Bergson is an exception) believe that our pres-
ent world was the inevitable outcome of forces that were
inherent in a fairly homogeneous molten mass, interact-
ing with those of its cosmic environment. It has never
been thought necessary to invoke the aid of special ‘‘de-
terminants’’ to account for the various geographic and
geologic features of our planet’s structure. In dealing
with inorganic things we are content to let our analysis
rest, in the lack of more detailed information, with the
acceptance of such general principles as ‘‘creative syn-
thesis’’ or the ‘‘multiplication of effects.” We simply

No. 627] ADAPTATION 353

have to admit that differentiation means just this fact of
de novo formation. Otherwise it means nothing at all.

We must, however, recognize certain essential differ-
ences between the development of a sea-urchin from an
egg and that of our world from the structureless spore
which was long ago liberated by its nebular parent. Let
us suppose that some experimental cosmogonist, using
the refined technique of a Morgan, Roux or Driesch, had
skilfully removed about three quarters of our newly
formed globe, leaving the remainder to reconstruct itself
as best it could. The spherical shape would doubtless
have been quickly restored, but is it likely that there
would have formed in the ensuing ages just that same
arrangement of Europe, Asia, Africa, America and the
Islands of the Sea that we now find upon our maps? Un-
fortunately it is too late to perform this experiment, but
I think that most geologists would expect a much modi-
fied world as the result. Indeed, if the excision had been
made after the mixture of molten substances had begun
to separate we should be perfectly certain that a quite
‘‘abnormal’’ world would have been the outcome. All
this may be granted.

Let us ask another question. Why is it that no modern
thinker?? has set forth a preformation theory of racial
evolution? It is only in accounting for individual devel-
opment that this has been thought necessary. Yet the
same paradox of de novo formation would seem to con-
front us in both cases, while other essential points of re-
semblance between phylogeny and = have often
been pointed out.

One difference, doubtless, is that every process of phy-
logeny is regarded as a unique thing, while ontogeny is
merely the nth reduplication of a known type, the char-
acter of which can be stated in advance. Hence it is that
we are satisfied to resign the former process to the realm
of ‘‘chance,’’ while the latter we come to look on as deter-
mined in advance. Another difference seems to be that
we look upon racial evolution as largely swayed by exter-

22 We must except Bateson.

354 ) THE AMERICAN NATURALIST [Vou LIII

nal factors, of the haphazard sort which operate in the
realms of geography and meteorology; while individual
development appears to be swayed chiefly by internal
factors, and to pursue its preordained course in a high
degree independent of the outside world.

But where in all this is the necessity for preformation?
That two specific types of protoplasm, under identical
conditions of environment, will give rise to widely differ-
ent organisms implies, of course, considerable difference
in the protoplasms. It does not, however, compel us to
believe in the existence of correspondingly numerous dif-
ferences in the two cases. A single initial difference be-
tween two physico-chemical systems may determine a
multitude of differences at the end. For example, the
presence or absence of a certain amount of annual rain-
fall on a given area of the earth’s surface would deter-
mine the nature of an indefinite number of other charac-
teristics, both geographical and biological. We do not
in this case endeavor to pick out a particular element of
the cause to account for each particular element in the
effect. Driesch’s assumption that any ‘‘mechanical’’
(i. e., non-vitalistic) conception of the developing organ-
ism must be based on a preformation of parts may once
more be dismissed as untenable.

Some preformation there is to be sure. Recent Men-
delian studies, particularly the investigations of sex de-
termination, make it highly probable that certain adult
characters, though perhaps in no case single anatomical
structures, are represented by spatially separated parti-
cles in the nucleus. Furthermore, a certain amount of
‘‘promorphology’’ has been demonstrated in the cyto-
plasm of the unfertilized egg, though this is perhaps to
be regarded as representing merely an early stage in
individual development. I feel bound to express the be-
lief, however, that many recent students of Mendelian
inheritance have carried their factorial speculations far
beyond the evidence, and that their detailed localization
of representative particles may prove in the future to
have more interest for psychology than for genetics. We

No. 627] ADAPTATION 355

are dealing with a field in which ever more minute differ-
ences are being distinguished—many of them by purely
subjective tests—and one in which the ratio of inference
to observed fact is ever lengthening. May it not be that
we have here hitherto unsuspected possibilities of self-
-` deception on the part of even our most competent inves-
tigators? The subject is one which seems to me to de-
serve more attention than it has received.

On the whole, we are not compelled to assume the
existence of any more preformation than can be experi-
mentally demonstrated. And it may be regarded as
settled that we have no parcelling out of ‘‘deter-
minants’’ to appropriate cells during ontogeny, such as
Weismann imagined. The ‘‘sex chromosomes,’’ which
seem to be the best authenticated instances of material
bearers of hereditary traits, do not pass into definite
body cells in the course of development and thus give
rise to the primary and secondary organs of sex. Rather
are they to be found distributed in every cell of the body.
The assumption that they set free their characteristic
determinants only in particular cells has no experimental
or observational foundation.

Now, I am quite aware that any such ‘‘intra-selection”’
hypothesis of organic regulation as has here been advo-
cated will be rejected by a large proportion of biologists
on the ground that it is entirely superfluous. Various
types of self-regulating mechanisms have been found in
the non-living world, and the phenomena of growth and
regeneration have long been known to be duplicated in
crystals. Przibram has gone to considerable lengths in
pointing out analogies between the behavior of the so-
called ‘‘fluid erystals’’ and that of a regenerating organ-
ism.2* And these analogies are reinforced by further
ones, based upon the regeneration of crystals of hemo-
globin. Many characteristically ‘‘vital’* phenomena were

23 (Archiv fiir Entwicklungsmechanik, October 16, 1906.) Likewise Tor-
rey (Scientific Monthly, December, 1915) has discussed some interesting

analogies between certain inorganic phenomena and the processes of ‘‘ accli-
matization’’ and ‘‘regulation.’’

356 THE AMERICAN NATURALIST [Vou. LIII

observed by him in these studies, among which the most
impressive was doubtless the making over of a softened
hemoglobin crystal by a process of ‘‘morphallaxis,’’ i. e.,
the readjustment of the matter already contained in the
fragment. There must thus be recognized in these non-
living masses of matter a tendency toward the attain-
ment of a specific form. And it seems plain that this
tendency may realize itself in more than one way. Yet
we should never, in this case, think of proposing any
hypothesis of ‘‘trial and error,’’ nor speak of the choice
by the crystal of ‘‘means”’ to an ‘‘end.’’

Now, I will hasten to express my own belief that the
phenomena in the two eases do not differ in any very
fundamental way. I am disposed to regard the regen-
eration of a crystal, the reconstruction of a mutilated
organism, and the solving of a problem by a mathemati-
cian as members of a single series of increasing com-
pleaity. They have in common the reattainment of a
condition of equilibrium which has been overthrown.
The fact that the organism is possessed of life, or that
the mathematician has a conscious end in view do not
alter the situation.

Such a ‘‘regulative’’ tendency in the inorganic world is
recognized by physical chemists as the ‘‘principle of
mobile equilibrium,’’ or the ‘‘theorem of Le Chatelier.’’
As stated by Lewis,?** this law asserts that ‘‘when a
factor determining the equilibrium of the system is
altered, the system tends to change in such a way as to
oppose and partially annul the alteration in the factor.
The same idea is conveyed by saying that every system
in equilibrium is conservative, or tends to remain un-
changed.” Bancroft?*» has given to this principle the
dignity of a ‘‘universal law,” pointing out analogies in
the realms of biology, sociology and economics. More
recently, its importance in ecology has been urged by
Adams.*”°

23a ‘(A System of Physical Chemistry,’’ Vol, II, 1916, pp. 140-141.

23b Science, Feb. 3, 1911.

23° AMERICAN NATURALIST, Oct—Nov., 1918; Jan—Feb., 1919. _

No. 627] ADAPTATION 357

In the regeneration of the more familiar type of
erystal, the latter doubtless goes about its task ‘‘un-
hesitatingly,’’ we may believe. But this is not true of
every inorganic system. ‘‘In a stream [of water],’’
says Jennings, ‘‘opposing actions of all sorts are com-
batted in ways almost as varied as in organisms: a
hole is filled up, a dam overflowed, an obstacle circum-
vented, another obstacle floated away, a bank of earth
undermined or cut through; and the stream finally
reaches the sea.’’24 Must we not recognize important
points of resemblance between such behavior and that of
a penned-up cat, scratching wildly at the objects in its
cage until finally a way out is found?

But if we admit this essential unity between the living
and the non-living in respect to their method of correct-
ing a disturbed equilibrium, why should we have resort
in one case more than the other to a theory of ‘‘contin-
gency’’ as regards the relation of means to end? Why
may we not suppose the regulative processes of proto-
plasm to proceed as directly toward a goal as those of a
crystal?

Answering the first question, I would say that the con-
ception of contingency has been introduced into this dis-
cussion merely in the sense of a denial of teleology.
Such a denial has been deemed necessary only in the case
of organic phenomena. For inorganic events are seldom
thought of as governed by ‘‘ends,’’ and the question of
‘‘means’’ does not therefore arise. But in this respect
there is really no difference between the living and, the
non-living.

The reason why the regulative processes of protoplasm
probably do not proceed as directly toward a goal as
those of a crystal lies, I believe, in the vastly greater
complexity of the former. But it does not seem likely
that any rigid distinction can be drawn. If it is really
true that a damaged crystal of hemoglobin can restore
its original form without the taking on of new material,
it seems hardly likely that this rearrangement is effected

24 Johns Hopkins University Circular, 1914, No. 10, p. 16.

358 THE AMERICAN NATURALIST (Vou. LIN

by the simple transfer of material from one point to an-
other along the straightest possible paths. There is
doubtless much random molecular movement which
serves only to retard the consummation of the process.

The more complex the system with which we are deal-
ing, the more of these ‘‘fortuitous’’ steps will intervene
between overthrow and recovery of equilibrium. The
chances that an entirely new disturbing factor will di-
rectly call forth the means to its own removal will corre-
spondingly decrease. -The morg plainly, therefore, will
the adjustment proceed in an ‘‘experimental’”’ fashion.”
Processes which favor the restoration of equilibrium (7.
e., Which satisfy the need) will be accelerated; those
which work in a contrary direction will be retarded.

At this point it may be profitable to cite certain closely
related utterances of Jennings :”°

The condition which results in . . . regulative action is the presence,
in a system, of a constant force, or stream of energy having a uniform
tendency or direction (or set of such forces), together with intermittent
forces having varied tendencies; whenever this condition exists, regu-
lative action appears. . . . When the constant stream of energy is re-
strained for some time from producing its usual effects, it overflows in
various directions, depending on the distribution of the resistance and
amount and intensity of the free energy. It thus produces one effect
after another. Often, at the end, one of these effects is of such a nature
as to overcome or avoid the restraint; the stream of energy may then
continue in the channel thus opened.

Has our prolonged discussion now led us, after all,
merely to a denial of the scientific validity of the adapta-
tion concept? I think not. The concept of adaptation
stands. upon the same footing as those of life, organiza-
tion, function, food, enemy, offspring, environment,
gece heredity and the scores of other indubitable
facts with which biology deals. By the use of pedantic
circumlocutions, all of these various expressions could
doubtless be avoided, and our ideas thus squared with
the most rigid demands of ‘‘mechanistic’’ philosophy.

25 Of course, such expressions as ‘‘experiment’’ and ‘‘trial and error’’
must be used in a strictly objective sense, so far as they are given any ex-

e.
26 Johns Hopkins University Circular, 1914, No. 10.

No. 627] ADAPTATION 359

But would such a renunciation bring us any nearer to the
truth? Only if we are ready to regard the whole science
of biology as a provisional one, a mere temporary rest-
ing place on the way to the more ‘‘exact’’ knowledge
which constitutes mathematical physics. How many of
us are prepared to make this admission?

Before passing on to the next subdivision of our field,
a few words are desirable in answer to another general
criticism which may be raised against the line of argu-
ment here followed. Exception may be taken to the ap-
parent assumption that the responses to a new situation,
whether physiological or psychological, are wholly. ran-
dom. Many responses are so obviously direct and un-
varying as to appear ‘‘fatally’’ determined.?*

Again, even where ‘‘experimentation’’ or ‘‘trial and
error’’is admittedly concerned in the process, the tenta-
tive efforts frequently lie within a quite restricted range
of possible movements, and from the first approximate
the goal to be reached much more nearly than if they
were wholly undirected. Thus the experiments of Hob-
house** upon various mammals suggest to him ‘‘that re-
cent writers have overestimated the effect of pure acci-
dent.’’ Furthermore, he concludes that ‘‘the more a
success was accidental the less likely were the animals
to take advantage of it.” So, too, in learning to throw
at a mark, we do not commence by casting our missiles
indifferently in every direction, but from the outset we
throw them in the general direction of the target. And
the same is palpably true when we attempt the solution
of a mental problem. The trains of thought are doubt-
less ‘‘spontaneous,’’ as pointed out above, but certain
more or less relevant trains are favored in advance. It
is from these that our selections are made.

Now, all these difficulties seem to me more apparent
than real. After the first dawn of conscious experience,
no situation is wholly new. Every problem which arises
contains elements in common with earlier ones which we

27 It is these which Loeb seems to regard as the more typical ones.
28‘*Mind in Evolution,’’ 1915, pp. 236-237.

360 THE AMERICAN NATURALIST (Von. LUI

have already solved. This is the more true the more
complex our problem. The ‘‘newness”’ of the latter may
relate to a very few features, the residue consisting of
elements which, in the last analysis, have been solved in
an entirely empirical fashion. And the same may doubt-
less be said of those adaptive physiological responses
which are generally assumed to be unconscious. As re-
gards the fixed reactions known as ‘‘tropisms,’’ I have
already pointed out the probability that the predomi-
nantly adaptive character of these has been the outcome
of racial history and therefore of some form of selection.

V. EVOLUTION AND ‘‘ContTINGENCY’’

In the two preceding sections of this paper stress has
been laid upon manifestations of the power of self-
adaptation in the individual organism. Very little has
been said regarding those fixed structural and functional
mechanisms by which the more usual needs of life are
provided for. The origin of such structures and func-
tions—‘‘adaptations,’’ as they are familiarly called—
must be accounted for in any adequate theory of evolu-
tion. Now, I have already argued that no theory of evo-
lution, so far as it is scientific, can admit the possibility
that the needs of the organism may call forth in any
direct way the initiation of those processes by which
these needs come to be satisfied. Let us look somewhat
further into this question.

The field of organic evolution is one which has lent
itself in a high degree to vitalistic and quasi-vitalistic
exploitation. From the time of the establishment of the
doctrine of descent, there were always persons who, in
spirit, still clung to the creation principle, while accept-
ing in form the newer ideas. Indeed, among biologists
themselves, there have always been those who have seen
in organic evolution the working out of a ‘‘perfecting
principle,” in a large degree independent of environ-
ment. Even Lamarck, who propounded one of the chief
naturalistic accounts of this process, admitted that life

No. 627] ADAPTATION 361

‘‘tends by its very nature to a higher organization.’’*®
The botanist Naegeli is one of the best known exponents
of such ai view. With some, like St. George Mivart, the
question has been closely interwoven with special theo-
logical beliefs.

This writer believed in an ‘‘innate tendency to deviate
at certain times and under certain conditions,’’ whic
tendency he held to be ‘‘an harmonious one, calculated to
simultaneously adjust the various parts of the organism
to their new relations.’’ And this guiding hand seems
to have been exercised not only in the direction of satis-
fying the needs of the organism itself, but in adapting
the latter to the needs of man. Speaking of the evolu-
tion of the horse, he tells us:

The series is an admirable example of successive modification in one
special direction along one beneficial line, and the teleologist must here
be allowed to consider that one motive of this modification (among
probably an indefinite number of motives inconceivable to us) was the
relationship in which the horse was to stand to the human inhabitants
of this planet.®°

Others, like Wallace, have had recourse to such a guid-
ing principle only in accounting for the origin of man.

In recent years, the philosopher Bergson has adopted
a vitalistic theory of evolution, weaving it into a meta-
physical system of which an important feature is the
essentially creative character of time or ‘‘duration.”’

e see the world of living things moving grandly on
through the ages, impelled by a mysterious force, the
‘‘élan vital,” and flowering out spontaneously into a
never-ending succession of living wonders. Such a con-
ception may stir the imagination, but it does not add to
our knowledge. $

Now, curiously enough, this ‘‘teleological’’ factor has
been introduced by various writers to explain two exactly
opposite classes of cases: (1) the origin of adaptive char-

29 Philosophie Zoologique (Elliot’s translation), p. 239, and elsewhere.
Lamarck’s statements are not wholly consistent, however, and I cannot feel
quite sure that he had in view any principle distinct from the one with
which his name is commonly associated

30 ‘* Genesis of Species,’’ p. 151.

362 THE AMERICAN NATURALIST [Von LIL

acters (Paley’s argument), and (2) the origin of highly
perfected structures and functions which are not be-
lieved to be adaptive in the biological sense, at least to
the extent of influencing survival. The musical and ar-
tistic faculties of man belong to this second class.

Natural selection, as is well known, provides us with at
least a formal explanation of the first class of characters,
but not of the second. Lamarckism, with a varying de-
gree of plausibility, accounts for the origin of characters
belonging to either class. That both of these theories
are, inlast analysis, theories of selection has been pointed
out in section II.

But the claim is to-day heard on various sides that

_. both natural selection and Lamarckism have broken

down completely, and that no other existing evolutionary
theories merit serious attention. _So impossible is it for
some biologists to square the widespread appearance of
adaptation in nature with their own special theories of
life that they seek to escape the dilemma by declaring
this appearance to be largely illusory. Thus Loeb?!
tells us:

While it is possible for forms with moderate disharmonies to survive,
those with gross disharmonies can not exist and we are not reminded of
their possible existence. As a ope: aay the cases of apparent adap-
tation prevail in nature.

In much the same vein, Davenport*? writes:

Strictly, we may say adaptation is not the thing that is brought about,
but rather absence of non-adaptedness. Such adjustment as we find is,
doubtless, only such a residuum of variants as has not proved incom-
patible with conditions of existence. :

One might profitably compare such conclusións as the
foregoing with the findings of Cannon,** based upon the
detailed study of certain adaptive mechanisms in man.
To most of us the conviction is doubtless irresistible, not

. that such mechanisms now exist because of their harm-
lessness, but that they came into existence, step by step,
on account of their utility.

31‘*The Organism as a Whole,’’ p. 344,

32 AMERICAN NATURALIST, August, 1916.

33 ‘í Bodily Changes in Pain, Hunger, Fear and Rage,’’ 1916.

No. 627] ADAPTATION 363

Taking heart from this skepticism among the biolo-
gists themselves, reactionaries are boldly coming for-
ward with the assertion that the evolution principle has
been discredited. It is certain that the spread of such
ideas is not calculated to further the advancement of
knowledge. Lack of an adequate hypothesis is not dis-
proof of any possible hypothesis.

Moreover, it would now seem that some of these ad-
missions of inadequacy have been premature. Much of
the recent abandonment of the natural selection theory
has been due to neo-Mendelian dogmatism. Selection, it
is claimed, can only separate strains having different
mean characters. It can not change the mean characters
of a pure strain. But the experiments of Castle and
some other breeders may be cited as evidence that such
a contention is far from being established. And even
those who reject Castle’s interpretation of these results
have been forced to concede that in some cases selection
may bring about the indefinite modification of our stock
—call the process ‘‘sorting’’ if we will.

So, too, the Lamarckian principle occupies the curious
position of being dogmatically denied or wholly ignored
by a large and influential class of writers, at the same
time that others are able to adduce apparently convince-
ing arguments for its reality. We certainly have a vast
array of indirect or circumstantial evidence for this prin-
ciple, derived from an inspection of the actual products
of evolution as we find them. And we have a certain
amount of direct, experimental evidence which can not
be thrown aside as irrelevant or untrustworthy. While,
therefore, sweeping conclusions regarding the Lamarck-
ian factor are doubtless premature, the dogmatic denial
of this factor very nearly amounts to self-stultification.

Thus, if we may read the signs of the times, the two
chief naturalistic explanations of evolution may survive
the fire of destructive criticism and again play an impor-
tant part in our interpretation of life. By this, I do not |
wish to be understood as arguing that either or both of
these theories constitute an adequate explanation (even

364 THE AMERICAN NATURALIST -[Vou. LII

in the sense of a description) of how evolution has come
to pass. For many years past, I have been endeavoring
to weigh the evidence for and against both of these hy-
potheses and I have reached the same verdict with re-
spect to the two: each is both proved and disproved. It
is not that adequate evidence is lacking, as some assume.
Rather, in each case, is the evidence well-nigh over-
whelming—on both sides.

Now, obviously, no single proposition can be both true
and untrue at the same time. What is meant here is
this. I believe the selection of virtually continuous vari-
ations and the inheritance of functional and environ-
mental modifications to have both played some part in
evolution. And I do not hesitate to say that the evi-
dence in favor of such a view is of the same general char-
acter as the evidence for the evolution theory itself, and
nearly as convincing.

On the other hand, it seems no less probable that the
operation of each of these factors is strictly limited. In-
deed, it would appear likely that much of the adaptive-
ness in nature is not adequately accounted for by either
process or by both taken together. There may well be
other factors the existence of which is as little suspected
to-day as was that of natural selection before the time of
Darwin and Wallace.

But will our explanations remain purely naturalistic,
or will they find room for extra-natural directive agents,
by whatever name called? Will they, like the two chief |
historic theories, base themselves on the contingency of
every adaptive variation in structure or function, ante-

_cedent to the test of experience, or will they be forced to
concede a primary adaptiveness inherent in living
matter.

Many of those who admit the widespread occurrence
of natural selection as a process, are wont to deny to it
any explanatory value. To quote a now familiar saying,
it is said that the survival of the fittest does not account
for the origin of fitness. The real cause of modification,
these writers insist, is to be sought in the process by

No. 627] ADAPTATION 365

which variations are produced and not in the fact that
many of these variations fail to maintain themselves.

This argument is so plausible that it seems self-
evident. And indeed in a sense it is. But there is an-
other sense in which it is quite specious. Truly enough,
no individual can survive which is not first born or
hatched, or in some way brought into being by its par-
ents. And those peculiarities which distinguish one in-
dividual from another are largely ushered into life along
with it. They exist prior to selection. But fitness is a
relation, not an absolute property of the organism. The
word denotes merely a certain measure of adjustment to
specific conditions of life, and the degree of this adjust-
ment we know to vary almost indefinitely. To say that
the conditions of life, acting through the selective proc-
ess, can not be the cause of an increasing degree of fitness
is like denying that a sculptor produces a statue, on the
ground that he does not create the stone. It is well to
note that even the sculptor’s function is wholly selective.
He eliminates certain portions of an unshaped mass of
material.*4

The foregoing analogy admittedly fails in one impor-
tapt respect. It implies that the possibilities of selec-
tion in a given race are wholly unlimited. We know this
to be very wide of the truth. The question to be an-
swered here is merely whether or not they are completely
random in the sense which has been employed through-
out this article.

Now, some selectionists are wont to deny the com-
pletely random character of variation. So far as this is
simply a denial of the infinite variability of any species,
it is a mere truism. We may perhaps admit the possi-
bility that a given strain might, through rigid selection,
acquire the ‘‘habit’’ of varying preponderantly in cer-
tain definite directions, thus limiting the possibilities of
further evolution within that group. And we might even
grant that such definitely directed variations might ac-

84T do not recall the previous use of this analogy, but it is such an
obvious one that it has doubtless occurred to many.

366 THE AMERICAN NATURALIST [Vou. LIII

cumulate without the influence of selection at all (ortho-
genesis). But can we, without departing from natural-
istic grounds, conceive of the production in this way of
a structure in anticipation of aneed? May we even con-
ceive how appropriate variations could be called forth
by an already existing need.

Of course, much obscurity of thought may be concealed
beneath this innocent-looking word ‘‘need.’’ What is a
need? It is notorious that what is a luxury to some of
us is a necessity to others. Our needs grow with our in-
comes. And this line of reasoning is directly applicable
to sub-human realms. What an animal has, if this ad-
justs it to certain conditions of the environment, may be
regarded retrospectively as the fulfilment of a need.
Thus eyes fulfil the need of seeing. But can we say that
such a need existed before the appearance of visual or-
gans? There are beyond doubt still many forms of
wave-motion or molecular vibration for which we have
no organs of perception. Thus, in a large measure the
organism creates its own needs, even in an unchanging
environment. The word ‘‘need,’’ like the word ‘‘end,”’
is one which has a distinctly teleological implication. The
more factors of the environmental complex an organism
is brought into relation with, the better is it adjusted to
its life conditions, and—other things equal—the higher
position it holds in the seale of life. But these adjust-
ments are only thought of as satisfying needs when we
come to look back on what has actually happened.*®

There is a more limited sense, however, in which the
use of this expression involves us: in no such obscurities.
All those fundamental requirements, such as food, oxy-
gen, protection from enemies, ete., may be termed needs,
without there resulting any confusion of thought. Now,
anything which. led to the removal of one or more of.
these fundamental requirements—say the drying up of a
lake—might bring about the extermination of an entire
species, unless some adaptive response were made.

35 They may all, however, be ‘properly termed adaptations, as has already
been said

No. 627] ADAPTATION 367

Here, likewise, we may legitimately speak of the need for
some sort of readjustment. Let us, then, restrict the
word to anything without which a species would become
extinct.

With this limitation of meaning understood, let us re-
turn to certain questions which I have left unanswered.
Can we, on naturalistic grounds, conceive how an appro-
priate trend of variation could anticipate a given need;
or can we even conceive how it could be called forth by an
existing need? The former possibility certainly can not
be admitted without frankly taking refuge in principles
which lie beyond the range of scientific analysis. The
latter possibility has, however, been vaguely implied by
some writers on evolution.

So far as the ‘‘need’’ might be the result of some
marked change in the environment or in the functional
activities of the organism, it is credible that new varia-
tions might be offered to selection as a consequence of
disturbances in the germinal material. But how could
these occur preponderatingly in the direction of meeting
the particular need in question? Only in one way, so far
as I can see, and that way is by the previous adaptive
modification of the parent body. For the latter may
adapt itself experimentally, according to principles al-
ready discussed. The germ-cells could not adapt them-
selves experimentally, since the need is commonly one
which does not as such affect them at all. Thus, the
imperative demand for directed germinal variations—or
at least ones of a useful sort—can be met, so far as now
appears, only by assuming the transmission to the germ-
cell of adaptive responses of the parent body.

The Lamarckian principle has the added advantage of
being able to account for many of the ‘‘luxuries’’ of
organization—adaptations, in the sense of fitting their
possessors for a fuller and more varied life, but not of
any conceivable survival value. Our own race, as has
often been pointed out, is endowed with multitudes of
such faculties. But we are sadly in need of direct ex-
perimental evidence along these lines.

368 THE AMERICAN NATURALIST [Von. LIII

Biologists of the future may recognize the importance -
of determining experimentally whether the germinal
variations of a species ever respond to changed life con-
ditions in such a way as to shift the mode of any char-
acter in the direction of greater adaptation. If sucha
general tendency as this were revealed, and if, at the
same time, the transmission of somatic modifications
were rigidly excluded, we should be brought to a crisis
in the history of our science. The question at issue
would not be merely the adequacy or this or that hypoth-
esis. It would be the adequacy of our recognized scientific
methods to deal with such problems. Despite the lengthy
arguments with which I have sought to defend a purely
naturalistic position, I should not, in advance, be su-
premely confident as to the outcome of such experiments.
It might, after all, turn out that there was just such an
‘immanent teleology’’ in living things as the vitalists
claim. If this should prove to be true, science would
have to re-survey its territory and set itself new bound-
aries well within the old ones.

Such an undertaking, like that of settling once for all
the ‘‘acquired characters’’ question, would doubtless be
beset by great technical difficulties. But these difficul-
ties should not be insuperable. So long, however, as
‘“genetics’’ is held to be nearly or quite synonymous with
Mendelism, evolution along dynamic lines is likely to lan-
guish. We must grant the enormous strides which have
been made in our knowledge of the inheritance of certain
types of variations, but the much more fundamental ques-
tion of the causes of these variations is almost as far
from solution as in the days of Darwin.

In conclusion, I would say a few further words in re-
gard to my use of the expressions ‘‘contingency’’ and
‘‘chance’’ throughout these pages. It is needless to say
that I have not used these words as synonymous with un-
caused. I have spoken of an event as contingent, merely
in the sense of its being causally unrelated to something
else: for example, a variation in relation to a need to be
fulfilled. Whether or not, in the last analysis, all things

No. 627] ADAPTATION 369

are causally related in an Absolute, or whether the Uni-
verse is pluralistic in its nature, need not concern us
here. That there may be some measure of pre-estab-
lished harmony among its various parts is possible. It
has recently been ably argued—and by a chemist, not a
theologian—that there exists such a pre-established har-
mony between the organic and the inorganic worlds as
a whole.*

But even granting such very problematic relationships
as this, we can not deny that much happens in a purely
‘accidental’? way. No degree of fitness on the part of
the environment for life in general can avail to prevent
the wholesale destruction of organisms which ‘‘happen’’
into unfavorable surroundings. That all of the special
adjustments between organism and environment arose
primarily through contingency or chance in the sense
here indicated is the main thesis which I have defended
in these pages. There may be little of an original na-
ture, either in the views proposed or the arguments used .
in support of them. But I believe that this essay may
serve a useful purpose in bringing together a number of
apparently distinct problems.under a common viewpoint.

86 L. J. Henderson: ‘‘The Fitness of the Environment’’ (1913), ‘‘ The
Order of Nature’’ (1917).

SHORTER ARTICLES AND DISCUSSION
PIEBALD RATS AND SELECTION, A CORRECTION

In a recent important publication Dr. Sturtevant makes ‘‘an
analysis of the effects of selection’’ in which he ably maintains
the current view that the single gene is not changed by processes
of systematic selection. His argument rests on a careful experi- -
mental study of the behavior of the character ‘‘dichaet’’ in Dro-
sophila, followed by a general discussion of other work, my own
in particular. I am represented as completely opposed to his
view, and so I have been at times, but such is not the case at
present. Iagree so fully with his general conclusion that I want
to obviate needless discussion based on the misapprehension.

I thought two years ago that I had evidence that a single gene
had changed in the course of a selection experiment, this gene
being concerned in producing the hooded pattern of rats. Inow
find this view rendered untenable by further experiments, the
results of which are in course of publication. These results
show that the supposed changes in a single gene are more prob-
ably due to changed residual heredity, which very likely may
consist wholly of other ‘‘modifying’’ genes.

he crucial experiment was one suggested by Dr. Sewall
Wright. The divergent hooded races, ‘‘plus’’ and ‘‘minus,’’ re-
sulting from selection, were to be crossed repeatedly with a third
race, the hooded character being recovered as a recessive in F,
following each cross and its variability compared with that of
the uncrossed race. It was believed that if multiple modifying
genes were involved, repeated crossing with a pure third race
would tend to remove these, in which case the extracted hooded
character being deprived of its plus modifiers would be sub-
stantially identical with the hooded character deprived of its
_minus modifiers, as seen respectively in hooded recessives derived
from the plus and from the minus crosses. Well, they are sub-
stantially identical, but it has taken some time and a good deal
of trouble to establish the fact. First we had to secure a satis-
factory third race to use in the crosses, one free from contami-
nation of any sort by crosses. This we sought in a wild race.
But ordinary wild rats will not breed under laboratory condi-
370

No. 627] SHORTER ARTICLES AND DISCUSSION Stl

tions. So we resorted to trapping immature wild rats from a
single locality and using these as a foundation stock. Crosses
with the plus race were then started successfully, but the corre-
sponding experiment with the minus race was hard to get going
and so has lagged behind the plus crosses. A report on the re-
sult of the plus crosses was made in 1916 (Castle and Wright).
The crosses with the minus race were not then sufficiently ad-
vanced to show what their outcome would be and this was still
true when reply was made to the criticism of MacDowell, as it
had been previously when reply was made to Muller. and to
Pearl, and subsequently when I addressed the Washington Acad-
emy of Science on the rôle of selection in evolution (1917). But
since then the minus crosses have given what seems to be con-
clusive evidence that the single gene had not been altered by
selection, although the inherited complex responsible for the

ooded character had steadily been altered in opposite directions
and these alterations were permanent in the sense that they rep-
resented racial modes, stable so long as the race was not out-
crossed.

I still have on hand a few representatives of the plus and of
the minus races which because of their low fecundity it has been
impossible to select further for several generations. The two
races are very different in appearance. The plus race shows no
white except on the under side and sometimes along the flank.
The minus race shows no black except a short hood lying anterior
to the shoulders, and in an occasional individual a small black
spot or two in the middle of the back or on the tail. Yet the
variability of each race is still considerable; as measured by our
‘“‘grades’’ it has not appreciably diminished in recent genera-
tions. The somatic differences entailed by the selection experi-
ments with the hooded character of rats are seemingly greater
than those secured by Sturtevant or by MacDowell in regard to
bristle number in Drosophila, yet I doubt not they may be ex-
plained on similar grounds.

Crossing with a wild race affects very differently the plus and
the minus selected races. See Tables I and II. The plus race
was much less affected than the minus race. Its mean grade was
lowered, by three successive crosses with the wild race, not over
three quarters of a grade. The standard deviation was about
doubled by the first cross. That is the variability of the hooded
character, when extracted in F, from the first wild cross, was

372 THE AMERICAN NATURALIST [Vor. LIL

about twice as great as the variability of the hooded character
in the uncrossed plus selected race. In the second and third
crosses the variability declined somewhat, but was still consid-
erably greater than that of the uncrossed race. It was indeed
very similar to that of the plus race in the first seven generations
of the plus selection experiment. (See Castle and Wright, p.
186.)
TABLE I

RESULTS OF REPEATEDLY CROSSING THE PLUS SELECTED RACE WITH A WILD
RACE

Number of
Mean Standard
Grade Deviation ee
Pana verges plus sy R eine + 3.73 36 776
Once extracted hooded Fz young............... + 3.17 a3
Twice Aa] hooded Fy. erp AL UNNA E + 3.34 50 256
Thrice extracted hooded Fz young.............. + 3.04 64 19
TABLE II
RESULTS OF REPEATEDLY CROSSING THE MINUS SELECTED RACE WITH A WILD
RACE P

N
Mean Standard Hooded
Grade Deviation Young

Control, uncrossed minus sea NEEE 16. — 2.63 Sri 1,980
Once extracted hooded Fz young............... — ,38 1,25 121
Twice extracted h F: eg PEE a DANE AGP AE + 1.01 .92 49
Thrice extracted hooded Fz young.............. + 2.55 66 104

The crosses of the minus race were started six generations
later in the course of the selection experiments, with animals of
generation 16, minus selection series. They show effects much
more striking than those of the plus crosses. See Table II. The
minus selected race had now attained a mean of —2.63. A
single cross, with the same wild race used in the crosses of the
plus series, lower the grade to —.38, extinguishing all the
changes in mean grade made by sixteen generations of selection,
and leaving the extracted hooded character in a highly variable
state (standard deviation 1.25, nearly five times what it had been
before). A second cross with the same wild race converted the
extracted hooded individuals for the most part into a plus group,
mean -+ 1.01, but with variability somewhat decreased, standard
deviation .92. A third cross with the wild race has given ex-

No. 627] SHORTER ARTICLES AND DISCUSSION 873

tracted hooded individuals exclusively plus in character, range
from + 1.00 to + 3.50, mean + 2.55. The variability has si-
multaneously fallen to .66, which is only about one third greater
than that of the minus race in the first five generations of the
selection experiment. (See Castle and Wright.) One family-
containing fourteen thrice extracted hooded individuals has a
mean grade for the hooded individuals of + 3.05, which is prac-
tically identical with the grade of the thrice extracted hooded
individuals resulting from the plus crosses (Table I).

It thus appears that three or at most four crosses with a wild
race suffice to obliterate all the racial differences which had been
induced by ten generations of selection in the case of the plus
race and sixteen generations in the case of the minus race. The
plus race was changed almost immediately by a single cross, but
the change was small (a fact which misled me until the results
of the minus were secured). The changes with the minus
race were so great that they could not be fully secured by less
than three or possibly four successive crosses (eight generations
of offspring). The wild race, which we used in our crosses, evi-
dently had a residual heredity much more like that of our plus-
selected than like that of our minus-selected race. When the
hooded gene from either race was introduced by repeated crosses
into this residual heredity, the result was to produce hooded
races of very similar grade, a little lower in grade than the plus
selected race, but very much higher in grade than the minus
selected race.

It thus becomes clear that the changes which had occurred in
the hooded character as a result of selection were detachable
changes and are probably in nature independently inherited
modifying factors. This is a view which Phillips and I gave as
one of two possible interpretations of the results which we pub-
lished in 1914. Morgan, Muller, MacDowell and others have in-
sisted that this was the only reasonable interpretation which
could be given, but I have not been satisfied with this conclusion
in advance of a really crucial experiment, such as I believe has
now been performed. Meanwhile the probability that the theory
of multiple modifying factors is correct as a general explanation
of similar cases has been greatly strengthened by the work of
Muller, Bridges, Sturtevant and others, showing that genetic fac-
tors, having a definite demonstrable position in linkage systems,
influence in a particular way the somatie manifestation of char-

374 THE AMERICAN NATURALIST [Vou, LIN

acters varying „quantitatively or qualitatively. I accept their ;
interpretations as correct in the light of our present knowledge.

I should feel like apologizing for my own obtuseness in not
reaching a similar conclusion sooner, did I not recall with satis-
faction how much clearer the réle of selection now stands re-
vealed than it did when these experiments were begun, and to
the clearing up of the situation I shall at least hope that this rat
work has contributed something, if only by provoking inves-
tigation.

The ‘‘Mutation Theory’’ of DeVries gave us a picture of se-
lection as an agency temporarily effective in producing racial
changes, but with those changes gradually vanishing as soon
as the selection ceased. Johannsen denied within ‘‘pure lines”?
even temporary effectiveness of selection. A strictly logical use
of Johannsen’s conclusions would have limited their application
to such organisms as he studied, self-fertilizing ones completely
homozygous for all genetic factors and subject apparently to
no new changes in such factors. But the doctrine was straight-
way extended in the views of most geneticists to selection of
every sort and he was treated as a traitor to Mendelism who saw
any utility in selection or advocated its use as a means of im-
proving the inherited characters of animals or plants.

The situation is wholly different to-day. Through the inves-
tigations of Jennings and his pupils on protozoa, of Stout on
Coleus, and of Shammel on citrous fruits, the fact is clear that
even within clones genetic changes may and do frequently occur
and that systematic selection will serve to isolate these and thus
lead to racial improvement. Those who have tried systematic
selection in the case of cross fertilizing organisms have in some
cases noted the occurrence of ‘‘mutations’’ with such frequency
as to make progressive change under selection easily obtainable.
Emerson and Hayes, in the case of certain pericarp color pat-
terns of maize, find ‘‘mutations’’ so common that a wide range
of variability results and selection is able to isolate, from such
material, types ‘‘relatively stable,’’ but very diverse in appear-
ance. Modifying factors are not involved in Emerson’s expla-
nation of his results, but rather such instability of a single gene
as leads to frequent mutation. Selection experiments with the
variegated coat-patterns of mammals seem to involve less abrupt
but otherwise similar changes, but modifying factors rather than
repeated mutation seems to be the explanation required in view
of the results of crosses reported in this paper.

No. 627] SHORTER ARTICLES AND DISCUSSION 375

That selection by one means or another is an effective agency
in producing racial changes is not questioned to-day, as it was ten
years ago.t The only question now at issue is whether the single
gene ischangeable. I am inclined to think, with Sturtevant, that
while single genes do occasionally change producing multiple al-
lelomorphs, a much more common occurrence is change in visible
characters through modifying factors. Whether the direction
of genetic variation is controllable, other than by the manipula-
tion of modifying genes or the discovery of multiple allelo-
morphs remains to be determined. The evidence at present is
largely negative. It is undeniable that liability to genetic varia-
tion is much greater in some organisms than others, much greater
as regards some kinds of character than as regards others, but
whether we can produce variability of a genetic character is
quite a different question. We certainly at present have to
follow nature’s lead rather than to lead nature, as regards the
course of evolutionary change.

W. E. CASTLE
BUSSEY INSTITUTION,
HARVARD UNIVERSITY.

1 Sturtevant’s presentation of my views is a bit unfair in that it seems
to imply that whenever I have spoken of ‘‘variation in a unit character,’’ I
have consistently meant variation in a single gene, whereas in discussing the
ease of the English rabbit, I have expressly reserved judgment on this point.
In a large part of my experimental work, the question under investigation
has been—do the visible characters which conform with Mendel’s law in
transmission suffer modification in crosses or as a result of selection? The
present opie of geneticists has nd geo! forgotten that this was
ever a debatable question. We all admit now that contamination occurs
in crosses va that modification may be mee by selection, an
only to explain how the contamination is brought about (as by modifying
factors) or how the modification is produced in the course of systematic

of ieee i i

the days when the doctrine of gametic purity was under discussion, such
‘*contamination’’ or ‘‘modification’’ was not admitted.

When Sturtevant denies the occurrence of ‘‘contamination,’’ he uses the
term in a very restricted sense, not as I have used it in a! Payne sen-
tence, nor as it was formerly used in BOPTI discuss: What he
means is not change in the visible charac as the iodid i cinchsion of
ae but change in a single gene which is Bnet absolutely to limit the

anifestation of the hooded character in any form. I agree with his view
rer there is no conclusive evidence that this single gene had changed in
the course of selection experiments, except in the case of our ‘‘mutant’’

376 THE AMERICAN NATURALIST [Vou. LII

BIBLIOGRAPHY
Castle, W.
1916, ae Selection Cause Genetic Change? AMER. Nart., 50.
ea Piebald Rats and Multiple Factors. AMER. NAT
1917a. let a le of Selection in Evolution. Jour. Wash. Keak Sci.,

» P-
Castle, W. co. aa “Hadle ey,
1915. The English hie imi the Question : Mendelian Unit-
racter Const yee c. Nat, Ac. Boi.,
Keer wie and Phi ities T
1914. i Spal Rats and Sictis Carnegie Inst. Wash., Publ. No.

195.

Castle, W. E., ‘and Wright, Sewall.

1916. Studios of Inheritance in Guinea-pigs and Rats. Carnegie

t. Wash., Publ. No. 241.

Emerson, R ER
1917. Genetical Studies of Vübpi Pericarp in Maize. Genetics, 2.
Hayes,

1917. Inheritance of a Mosaic Pericarp Pattern Color of Maize.

Dee netics, 2

- Jennings, H. S.
1916. uot Variation, and the Results of me ey in Uniparental
Reproduction in Difflugia corona. Genetics, 1
MacDowell, E. C.
1916. Piebald Rats and Multiple Factors. AMER. Nar., 50.
1917. Bristle Inheritance in T Jour, Exp. Zool, 23.
Stout, A. B
1915, The Establishment of Varicties in Coleus by the pug: of
Somatic Variations. Carnegie Inst. Wash., Publ. No.
Sturtevant, A. H.
1918. An Sgr ow of the Effects of Selection. Carnegie Inst. Wash.,
Publ. No. 264.

THE
AMERICAN NATURALIST

Vor. LILI. September—October, 1919 No. 628

EGG-WEIGHT AS A CRITERION OF NUMERICAL
PRODUCTION IN THE DOMESTIC FOWL!

DR. PHILIP HADLEY

I

Wuen one surveys the field of literature dedicated to
the subject of egg-production in the domestic fowl he may
well be astonished at the vast number of ways and means
by which a poultryman can detect the best layers of his
flock. Indeed one exaggerates only slightly to say that
there is scarcely an incident in the hen’s daily program,
nor an event in her life, that has not been interpreted by
some unusually keen observer as a sign of producing
ability—good or poor. Was the hen seen to rise early and
dispatch a one-hundred-calory portion of mash, together
with nine bugs and three worms, before her sisters were
off the roost? Then put her down unqualifiedly as an
industrious hen and enthusiastic layer—a credit to any
poultry house. Was the hen observed to work after hours
gleaning the last fragment of grain from the litter, or
perchance chasing lightning bugs through the twilight
grass, when other union-members of the flock had ceased
work for the day and retired to roost? Then register her
as one that has her master’s interests at heart, and one
that should be vigorously encouraged to reproduce her
like. Did the hen lose the yellow glamour of her shanks
and beak (doubtless the equivalent of good complexion in
ahen)? Did she molt in July or August? Was her comb

1 Contribution 250 from the Agricultural Experiment Station of the Rhode
Island State College.

377

378 THE AMERICAN NATURALIST [ Vou. LIII

a ruddy red in September? Was her pelvis broad and
flexible in April? Did she start laying in October? Did
she lay thirty or more eggs before the first day of March?
Did she lay 200 eggs in her pullet year, or 500 eggs in
three years? Did she lay small eggs or large eggs? By
all these signs one may (it is alleged) detect the hen that
is (or has been) the good producer. But the curious part
of the matter is that, notwithstanding these many signs
and evidences of producing ability, the hens of the aver-
age poultryman continue to deliver the same number of
eggs per year—estimated at about 120.

Among this variety of criteria, however, it must in
fairness be said that some of the tests are of practical
significance. It can scarcely be doubted that, as a rule,
hens that lay the largest number of eggs during the
‘winter period’’ (November 1 to March 1), as first stated
by Pearl, are the best layers for the entire year. On the
other hand, it has been shown by Goodale that the produc-
tion during the winter period may be strongly influenced
by the time of hatching: the early-hatched hens make the
highest winter records—at least they lay the greater
number of eggs between the beginning of the laying period
(sometimes as early as August) and March 1.?

If a hen is entitled to be called a good producer only on
condition that she makes a creditable record for two or
more years successively—then there is point to the recent
contention of other investigators that hens that make a
low first year’s record usually ‘‘make up’’ during the
second year, so that a three-year production record ap-
pears to them as representing the fairest measure of pro-
ducing ability. This is of course the equivalent of saying
that the number of eggs that a hen lays is a good criterion
of her egg-producing ability—a circumstance which no
one can deny. But it frequently happens that, for prac-
tical purposes, one desires such a criterion as will indi-

2It may be a question, however, whether the ‘‘ winter period’? of Good-
ale’s early-hatched pullets may not in reality represent a combining of two

laying cycles. His data on production seem to make possible this interpre-
tation

No. 628] EGG-WEIGHT 379

cate a hen’s producing ability before she has attained that
stage in life when economic production ceases; and when,
even as a breeder, her further producing days are few.

II

To the casual reader it will no doubt appear preposter-
ous that a biologist should attempt to measure the numer-
ical egg-production of a hen by weighing her eggs, rather
than by counting them. But the author freely admits
that this ridiculous thing has actually been done in his
laboratories; and, what is more, that the method appears
to work: a flock of hens can be divided into groups, each
characterized by a different mean producing ability, as a
result of weighing a certain number of eggs at a certain
time in the laying year, and subsequently by making cer-
tain computations therefrom. The results depend upon
the relation existing between egg-weight and egg-produc-
tion at different periods of the laying year. These points
may be considered separately.

deg tt: San Yeo Mar Ayr Mau June July Aua Sept Get Nev Dee Jan “Fete Mar Apr May Tune
‘
a
Ton beheld Man.
‘ Ef
$ ww ma z E
4 “x
3 Seo a3
? gz
2 be
$ an 3 + $s
s Joo £
"i E [$9
H ‘ |
E Reo A E EE om ARN
z ni |
3 2 |
RE SERAN
5 TERA et
aN 2 Tees es
š we oe
-
È ow {pa S S Bat---4
b
£ =e Sate,
= E
=
ba P F 4 ži
k fn 4 L4
> R-
A S A AEE E r---4 cx
s ys
P St
ił Se - Savaa Seana ia peg
“Riese Saying Xear{(taniere) * (parria)
Fig. 1.

When the first yearly production of a flock of hens of
equal age and condition is plotted on monthly ordinates
one frequently obtains a curve such as indicated by the
full line in Fig. 1. It is observed that the production

380 THE AMERICAN NATURALIST [ Vou; LIH

curve for the year ending October 31 is of the bimodal
type.®

One of these modes, appearing on the spring of the
year (in April), may be termed the vernal production
maximum. The second mode, falling in September, may
be called the autumnal production maximum. Of these
two maxima for the first laying year the vernal maximum
manifests the higher peak, in keeping with the heavy
April production which is the highest of the year. Be-
tween any two successive maxima lies a fundus whose
minimum is attained either in July or August, or in No-
vember of each year. The late summer depression may
be termed the estival production minimum, and the No-
vember depression, the autumnal production minimum.
In the month of December of the second laying year it
will be noted that the curve rises slightly. This repre-
sents the increased production of the ‘‘winter cycle’’ or,
as it may be called, the hibernal production maximum.
Between this and the vernal production maximum of the
second laying year is another depression in the curve,
following the period of winter production and indicating
the hibernal production minimum. This is followed in
turn by the vernal maximum of the second year.

The presence of these two maxima in the curve of pro-
duction for the first laying year indicates that at least
twice during this year (closing October 31), once in the
spring and once in the autumn, the egg production of a
hen rises from its lower limits and expresses itself by
laying a larger number of eggs than at any other season.
These periods of increased production represent the
spring and the autumn cycles respectively. There exists
also the winter cycle which is usually manifested with
clearness only in those flocks which show a fairly high
- mean production. It is a significant productive period,
but will not receive further consideration at this time.
It may be added, however, that in birds that are fairly
high producers and which are hatched sufficiently early

8 These hens were poor winter producers and the expected mode for De-
cember-January of the pullet year does not appear.

No. 628] EGG-WEIGHT 381

in the year, the winter production maximum may make
its appearance as the first cycle of production of the first
laying year.

We have considered some of the more obvious varia-
tions in the curve of numerical production and come now
to the curve of variations in egg-weight. Such variations
may be considered with reference to the clutch, the litter,
the cycle or the year.*

For present convenience, however, we shall consider
the variation by months—purely arbitrary divisions in
the life of the hen, which cut in on, and interrupt
clutches, litters and gelea: in such a way as frequently to
obscure many of the problems involved. For our present
purpose, however, division by months offers a rough and
ready division of the year into short periods of time in
which the productions may be compared.

When all the eggs laid by a flock of hens are weighed
and recorded, and the monthly means computed and
plotted on monthly ordinates, such a curve of mean
monthly variation in egg-weight is obtained as that
shown by the broken line in Fig. 1. Such a curve shows
that all the eggs that a hen lays are by no means of equal
weight. The first eggs laid are relatively small, but the
weight increases gradually until a maximum weight for
the first year is attained in the month of April. This is
termed the vernal weight maximum and may represent
mean increase of five grams over the mean weight of eggs
for the first laying-month of the year. This maximum
forms the first mode of the frequency curve of variation
in egg weight as shown in the figure.

-After April, the curve of variation in egg-weight drops
for May, again for June, and reaches the lowest point in
July, at which time the mean weight of the eggs of the
flock may be scarcely greater than for the first month of
production. Having struck this low point, however, the

4A clutch may be regarded as the group of eggs laid on successive days
without an interruption. A litter is the group of eggs laid immediately

preceding the onset of a broody period. A cycle is the larger group of eggs
laid during any one of the seasonal periods of increased production.

382 THE AMERICAN NATURALIST [ Vou. LIII

mean egg-weight begins to rise again and reaches a sec-
ond mode or maximum in September, at which peak the
mean egg weight is slightly higher than for the period of
the vernal maximum (April). This September peak may
be called the autumnal weight maximum. The mean dif-
ference between the vernal maximum and the autumnal
maximum is usually about one or two grams. Having
attained this peak of weight, the curve drops again
through October to strike its fundus in November (the
first month of the second laying-year).° From this point
it rises in December to the first weight maximum
(hibernal weight maximum) of the second laying year,
and then drops again in January to form the hibernal
weight minimum immediately preceding the vernal max-
imum of the second year.

It will now be clear to the reader that there exists a
noteworthy circumstance with reference to these curves
of numerical production and of egg-weight: they parallel
one another to a remarkable degree. The vernal maxima
of production and of weight fall together in April; and
the autumnal maxima of production and of weight fall
together in September. The only departure from coin-
cidence lies in the circumstance that the summer produc-
tion minimum arrives in August, while the summer
weight minimum is found in July. It should be said,
however, that the plotting of the curves on ten or five-day
_ ordinates might show a closer correspondence of these
minima in point of time. The difference observed is
-scarcely significant. The definite agreements in the
trends of the respective curves are taken to indicate that,
on the average, increased production is accompanied by
increased mean weight of the eggs produced; and that,
vice versa, a decrease in production is, on the average,
accompanied by decreased mean weight in the eggs pro-
duced. Whatever, therefore, may be the biological sig-
nificance of the two production maxima for the hen’s first

5 It has become common to consider the laying year of a hen as extending

from November 1 of the pullet year to and including October 31 of the year
following.

No. 628] EGG-WEIGHT 383

laying year, the weight maxima would appear to possess
a similar significance. Since the two are so closely cor-
related it would seem possible to measure a hen’s innate
egg-producing ability by the one phenomenon as well as
by the other. This constitutes the hypothesis which we
will now attempt to verify.

IT

If we take a cross-section of the April production as
nearly as possible to the absolute mode® of the weight
curve, we learn that, although the egg-weight of most of
the individuals of the flock has increased at this time,
there are a few in which it has not increased significantly ;
and a still smaller number in which there has occurred a
loss in egg-weight. The same is true for a cross-section
of production taken at or near the absolute mode of the
autumnal weight maximum. The following question
therefore arises: Does there exist any significant corre-
lation between a tendency to manifest an increase in egg-
weight at the period of the vernal weight maximum (or
autumnal weight maximum) and the number of eggs pro-
duced for the entire first laying year (November 1 to
October 31 following)?

In order to demonstrate such a correlation one must
first define more exactly the nature of the second vari-
able, namely, the ‘‘tendency to manifest increased egg-
weight’’ as referred to above. There must be a fixed
point from which one can calculate, for each individual
hen, the amount or the extent of increase in egg-weight
manifested at the weight maxima. For certain reasons

explanation of this point it may be added that by plotting the fre-

would vary with different flocks, depending upon the climate, the date of
hatching, the method of housing and presumably upon still other varying,
environmental factors.

384 THE AMERICAN NATURALIST (vorn LIII

it was decided to compute all increase or decrease in egg-
weight, for each individual, from the mean weight of the
first ten eggs laid at the beginning of the first laying year
of that particular bird. And, in order to translate the
differences into comparable terms, the increase or de-
crease was calculated as a percentage increase or as a
percentage decrease above or below the mean weight of
those first ten eggs. ‘Consequently, the percentage of in-
crease or decrease in mean weight for all April eggs,
over or under the mean weight of the first ten eggs laid,
was ascertained in the case of each bird in the flock; and
the same data were derived for the September produc-
tion. It is upon the analysis of these raw data‘ that the
appended computations rest. In the succeeding para-
graphs it is therefore the aim of the writer to demon-
strate the following point: that the higher percentages of
increase in mean egg-weight, reckoned at the periods of
the weight maxima, are so closely correlated with higher
production for the first laying year, that, by the method
to be presented, a flock of hens may be divided into
groups characterized respectively by higher, medium and
lower producing ability ; and that this method is effective,
whether the computations are based upon the vernal or
the autumnal weight maxima.

IV

We may first concern ourselves with computations
based upon the mean weight of the April eggs, including
the eggs of the entire month; and it is scarcely necessary
to resort to formal correlation tables to demonstrate the
point involved. ‘The simpler methods may be employed:
(1) Dividing the birds on the basis of annual production
above or below the flock-mean and then computing the
percentage of net increase or decrease in mean egg-
weight; (2) dividing the birds into groups based upon
percentages of net increase or decrease in mean egg-

7 It would be impossible to present these raw data in an article of this

They will be published, however, at the close of the investigation
which is still in progress.

No. 628] EGG-WEIGHT 385

weight and then ascertaining the mean annual produc-
tion for each percentage-group. To make the matter
more clear both methods will now be applied—first to a
differentiation of the flock on the basis of production
groups.

For the purposes of the present inquiry the flocks may
be divided into two groups on the basis of the mean
annual production which was 120 eggs. One group was
made up of individuals whose production was above the
mean, and the other group included birds whose produc-
tion was below. The mean production of the plus group
was found to be 143 eggs, while the mean production of
the minus group was 99. After these production-groups

TABLE I
SHOWING THE PRODUCTION OF THE First LAYING YEAR OF GROUPS OF BIRDS
CTED FOR DIFFERENT PERCENTAGES OF INCREASE OR DECREASE IN
MEAN Ece WEIGHT, MEASURED AT THE PERIOD OF THE VERNAL
(APRIL) WEIGHT MAXIMUM

1 Dirin Sileoeat tor | Number of | Mean Pro- P Sirds Selected for | Number of | Mean Pro-
Increase in Egg Individuals | duction for Increase in Egg Individuals | duction for
Weight as Indi- | Making the t Weight as Indi- | Making the | the First

cated Below Record Laying Ycar cated Below Record La: Y
> 10 per cent 2 147 > 4 per cent. 17 132
> 9 si = 3 142 > 3 “ ee 21 122
Se Bn eas 4 141 O Ea 31 122
oa z oe oe 6 140 < 0 of oe 6 1 1 1

E i a aly 9 137 Tee 28 112
a E AE PG 16 134 |

Total flock... 379 120

had been established the mean net increase or decrease
of egg-weight for each group was computed. The mean
increase for the plus group was 5.4 per cent., and for the
minus group 2.6 per cent. These results appear to indi-
cate that, on the average, birds which manifest a greater
percentage of increase in the weight of April eggs are
likely to be the better producers of the flock.

In utilizing the second method mentioned above, the

8 The flock in question consisted of 38 white Plymouth Rocks hatched in
April, 1909. Some of the birds have now completed their seventh laying

year.
9 One hen, showing no increase and no decrease in mean egg-weight, and
a production of 91 eggs, was omitted from the records.

386 THE AMERICAN NATURALIST [ Vou. LIII

birds of the original flock were divided into groups ac-
cording to the percentage of net increase (or decrease)
in the mean weight of eggs laid during the period of the
vernal maximum (April). The percentage-groups were
based on the scale indicated in the accompanying table.

From the data presented in Table I it is apparent that,
on the average, the birds that showed the higher per-
centages of increase in the weight of the April eggs were
also characterized by the higher productions. Those
characterized by a weight-increase of 10 per cent. or
more showed a mean production of 147 eggs, while those
characterized by a weight-increase of more than 3 per
cent. only, showed a mean production of only 122 eggs.
The mean production of the group characterized by a .
decrease in egg-weight was the lowest of all—111 eggs,
this being below the mean production of the entire flock.

If the birds are divided into two groups only, one
having an increase of 6 per cent. or more, the other show-
ing an increase of less than 6 per cent. or an actual
decrease in egg-weight, it is found that the high-per-
centage group gives a mean production of 137 eggs, while
the low-percentage group gives a production of only 112.
In this instance the portion of the flock falling in the
high-percentage class was approximately 24 per cent. ;
and this small group gave an average production that
was 23 per cent. in excess of the production of the low-
percentage group. The fact is thus brought out that,
although a certain small proportion of high-producing
individuals that are also characterized as manifesting
only a slight percentage of increase in egg-weight at the
period of the vernal weight maximum, will usually be
found, the higher producers are, on the average, char-
acterized by the larger percentages of increase (6 per
cent. or above); and the selection of hens on this basis
results in the separation of those individuals possessing
the highest producing value.

No. 628] EGG-WEIGHT 387

y

In view of this correlation between numerical produc-
tion and percentage of increase in egg-weight when meas-
ured at the period of the vernal maximum, it appeared
desirable to ascertain whether a similar correlation ex-
isted between production and increase in egg-weight
manifested at the autumnal (September) maximum. The
same two demonstrational methods used in the previous
instance may be applied.

The data on production were first re-distributed in
such a manner as to group the percentages of increase or
of decrease in egg-weight under two headings: (1) hens
having an individual annual production greater than the
mean (120 eggs), and (2) hens having an individual
annual proportion of less than the mean production of
the entire flock. In this way it was brought out that the
plus group, with a mean production of 151 eggs, showed
a mean net increase in egg-weight for September of 5.8
per cent., while the minus group with a mean production
of 105 eggs showed a mean net increase of only one per
cent.

TABLE II
SHOWING THE PRODUCTION oF THE First LAYING YEAR OR GROUPS oF BIRDS
SELECTED FOR DIFFERENT PERCENTAGES OF INCREASE OR DECREASE IN

MEAN EGG-WEIGHT, MEASURED AT THE PERIOD OF THE AUTUMN
(SEPTEMBER) WEIGHT MAXIMUM,

Puncunrace-cuass (“Number of | Mean Pro- || Purcewracn-cuass| Number of | Mean Pro-
Birds Selected for | Individuals | duction for Bi Selected for Individuals duction for
Increase in Weight | Making the the First Increase in Weight | Making the the First
Indicated Below Record Laying Year| Indicated Below Record Laying Year
> 10 per cent. 8 143 >4 per cent. —
ia | acne 9 141 >3 18 131
E 10 140 >0 y 25 127
bes e Sic Maui 11 142 itt een os 8 108
bh, r 12 141 ea TARA TE EIA 21 111
a. P 14 137 a
Total flock. .. 3310 120

When the second method was applied, and the data re-
distributed so as to give the percentage-classes, the re-
sults shown in Table II were obtained.

10 Four hens included in Table I were not employed in the present com-
putations, because they failed to lay during September.

388 THE AMERICAN NATURALIST [ Von. LIII

From the data presented in Table II it appears that, on
the average, the birds that manifested the higher per-
centages of increase in the weight of the September eggs
were characterized by higher annual production. Those
showing a weight-increase of 10 per cent. or more gave a
mean annual production of 143 eggs, while those birds
characterized by a decrease in mean egg-weight. showed
an annual production of only 108 eggs. When the flock
was divided into two groups according as the egg-weight
had increased by more than 6 per cent. or less, the high-
percentage group gave a production of 141 eggs as op-
posed to 111 eggs laid by the low-percentage group. Thus,
dividing the flock on the basis of a 6 per cent. increase in
the mean weight of all the September eggs, gave a group
of 12 hens (out of 33) which showed a mean production
17.5 per cent. higher than the flock average (120), about
27 per cent. higher than the mean production of the low-
percentage group, and 30 per cent. higher than the mean
production of the small group of eight hens which mani-
fested a decrease in mean egg-weight at the period con-
sidered.

It will hardly be necessary to call the attention of the
reader to the circumstance that this method of demon-
strating the correlation involved in the frequency distri-
bution of these two variables (increase in egg-weight and
numerical production) is, by its very nature, such as to
constitute a practical application of the means involved.

The correlations between weight-increase and produc-
tion, considered in the foregoing paragraphs, were so
obvious that the question arose as to whether satisfactory
correlations could not be demonstrated between these two
variables under conditions in which a smaller amount of
statistical data was employed. For instance, if the
method should prove of value to poultrymen in affording
a means for the detection of the higher producers of the
flock, it would be desirable to reduce the machinery of
computation to the lowest point consistent with valid
results. ` It thus appeared pertinent to inquire whether
computations based upon the weight of only ten eggs, laid
as closely as possible to the periods of the absolute vernal

No. 628] EGG-WEIGHT | 389

and autumnal maxima, respectively, would afford a satis-
factory basis for establishing the weight-production cor-
relations.

To this end, therefore, the mean weight of ten eggs laid
by each member of the flock between the eleventh and
twenty-fifth days of April't was computed, and the differ-
ence between the mean weight of these ten eggs and the
mean weight of the first ten eggs laid by that hen at the
beginning of her laying performance calculated as a per-
centage-increase or as a percentage-decrease. It should
be added that the production during April was conducted
at so rapid a rate that, in the case of 28 individuals out of
37, it was possible to obtain the record of ten eggs within
the dates mentioned. In the remainder of individuals it
was necessary to transcend these limits slightly. In no
instance, however, was it necessary to take eggs from a
date earlier than April 8, nor later than April 29. The
data thus acquired were redistributed according to the
percentage groups, and the results summarized in Table

TABLE III
SHOWING THE MEAN ANNUAL PRODUCTION FOR THE First LAYING YEAR OF
GROUPS OF oe SELECTED FOR VARYING PERCENTAGES OF IN-
CREASE OR DECREASE IN EGG-WEIGHT, COMPUTED ON THE BASIS OF
THE WEIGHT OF os Eces LAID AT THE ee OF THE
VERNAL WEIGHT MAXIM

PERCENTAGE-CLASS:| Number of | Mean Pro- |[PERCENTAGR-CLASS: Number of | Mean Pro-
Birds Selected for | Individuals | duction for || Birds Selected for | Individuais | duction for
Increase in Weight | Ma the the First || Increase in Weight | Making the | e First

Indicated Below Record Laying Year | Indicated Below Record | Laying Year
> a per cent. 3 142 > > r per cent. 27 | 123

> 7 141 29 | 122

Ss > tt u 9 140 || a o t t 31 | 120
Bean: E 13 138 | wh) e ety 16 | 106

Dal Sha E 19 WE ee o A 7 I

eee Sar p 22 126 || |

oe 23 125 |i Total flock...' 37 oort

Mein the data presented in Table ITI it is clear that the
small group of hens characterized by a percentage-
increase on egg-weight of more than 10 gave a higher
mean production (142 eggs) than any group manifesting a
smaller percentage of increase in egg-weight. Each suc-

11 See footnote on page 383.

390 THE AMERICAN NATURALIST [Vou. LIII

ceeding group, characterized on a smaller percentage-
increase, gave a correspondingly smaller annual produc-
tion, until, when we reach ‘‘<0 per cent., the group
manifesting a decrease in mean egg-weight, we find a
mean annual production of only 106 eggs. When the
flock is divided according as the mean percentage of in-
crease is more than 6, or less than 6, we find that in the
high-percentage group there are 13 hens with a mean
annual production of 138 eggs, while in the low-percent-
age group there are 24 hens with a mean production of
114 eggs. In other words, upon the division point of 6
per cent. increase, one may separate about one third of
the flock whose annual production is 15 per cent. higher
than the flock average and 21 per cent. higher than the
mean production of the remainder of the birds.

If the reader will now make a comparison of the results
reported in Tables I and ITI, it will be seen that the corre-
lation demonstrated through the employment of the ‘‘ten-
egg method’’ is as clearly established, and as valuable
from the practical point of view, as the correlation demon-
strated through the use of a full month’s production.

VII

In view, therefore, of these results obtained from the
weighing of ten eggs at the period of the vernal weight
maximum, it seemed desirable to ascertain whether the
` same ‘‘ten-egg method’’ at the period of the autumnal
weight maximum would also serve to distinguish a group
of hens characterized by the possession of higher pro-
ducing ability. Accordingly the production data for
September were analyzed from this point of view.

In explanation of the September results, however, sev-
eral points should be noted. In the first place, although
September production represents a definite mode in the
annual production curve when plotted on monthly ordi-
nates, in the case of the flock studied the month’s pro-
duction falls considerably short of the April production.
In April all members of the flock, without an exception,
were laying. In September there were four hens that did
not lay at all; and three hens laid only three eggs or less.

No. 628] EGG-WEIGHT 391

In the redistribution of the data for the present purpose
the records of no hens are included that did not lay at
least five eggs in September. Two hens laid nine eggs
five eggs. So that, in reality, the results of this case are
based upon the mean weight of somewhat less than ten
eggs from each hen.

In the second place it should be noted that the Septem-
ber production was scattered when compared with the
April production; and although an attempt was made to
secure eggs laid during the latter half of the month, it
frequently happened that it was necessary to include eggs
laid’ in the earlier part. The results of this redistribution
of data and the attendant computations are presented in
Table V.

TABLE IV
SHOWING THE MEAN ANNUAL PRODUCTION FOR THE FIRST LAYING YEAR OF
GROUPS OF INDIVIDUALS SELECTED FOR VARYING peenes oF IN-
CREASE OR DECREASE IN EAN GG- WEIGHT, COMPUTED ON THE
BASIS OF THE WEIGHT OF TEN EGGs OR LESS, as AT THE
PERIOD OF THE AUTUMNAL WEIGHT MAXIMUM.

PERCENT. ASS:| Number Mean Pro- PERCENTAGE-CLASS: Number of ean Pro-
Birds Selec at ror Individuals duction for Birds Selected for Individuals Santis for
Increase in Weight | Making thi the First Increase in Weight Making the, the First
Indicated Below Record rite Year Indicated Below Record (Laying Year
> 18 per cent. 3 147 E per cent. 21 125
be oe A i 6 145 ae | 23 125
po cae (0 Dineen 7 143 >0 Z 26 124
lene” E 8 44 R sks 5 108
ph R ial ek 10 1 EE S 19 112
e TY = adp TOA 2 a
Ba 12 139 f
Sag 4s 14 is es ee d 6
sa en i 16 i34 | Ditto „plus sta vel
Ss 9 te 19 131 ing ponies in tie 12 101
Total fi s 31 120

From the data presented in Table IV it appears, as in
the former case, that higher production is correlated with
the higher percentages of increase in egg-weight. The
maximum group-production (147) occurred in those hens
whose mean increase in weight was above 13 per cent.
Selecting above 10 per cent. gave seven birds whose mean
production was 143 eggs. Selecting above 6 per cent.
gave twelve hens whose mean production was 139 eggs.
On the other hand, selecting below 0 per cent. (i. e., birds

392 THE AMERICAN NATURALIST [ Vou. LIII

showing a decrease in egg-weight) gave five hens with a
mean: production of only 108. When we add to these the
hens that laid three eggs or less in September, we obtain
a group whose mean production was only 96; and when
we consider the hens that (1) gave a September produc-
tion of 3 eggs or less, and (2) gave a decrease in egg
weight, we obtain a combined group of 12 whose mean
annual production was only 101 eggs for the first laying
year.

A comparison of Tables I, II, III and IV thus shows
that the last case presents the clearest evidence yet ob-
tained for the positive correlation existing between per-
centage of increase in egg-weight and total annual pro-
duction. The results are more definite than those ob-
tained for the ‘‘ten-egg test’’ at the vernal weight maxi-
mum, or for the ‘‘month test’’ at either the vernal or the
autumnal weight maxima. In other words a test based
upon a smaller number of eggs, laid nearer to the absolute
mode, gives a clearer indication of innate producing abil-
ity than does a test based upon a larger number of eggs
laid in a ‘‘seatter grouping’’ about the approximate mode.
This conclusion is in harmony with the views expressed
by Gavin’ and by Wilson’ to the effect that the best unit
of time for measuring a cow’s milk-producing ability is
not the year test, nor the thirty-day test, nor even the
seven-day test, but the one-day test conducted when the
production reaches its maximum. Apparently the meas-
urement of egg-production in the domestic fowl, consid-
ered as a ee performance, rests upon a similar `
basis.

Vil

In bringing this paper to a close the writer wishes to
have it distinctly understood that nowhere in these pages
has it been stated that there exists in the domestic fowl a

12 Jour, Royal Agricultural Society, 1913, 73. Jour, adtag Society,
1913, 5, 309-319. Ibid., 1913, 5, 377-390 (on authority of Pearl).

18 Proc. Royal Dublin Bostely, 1911, 13, 89-113. Jour. back: Agricul-
ture, Ireland, 1913, 13, (4) (on aathiorite of Pearl).

No. 628] EGG-WEIGHT 393

correlation between egg-production and egg-weight. Most
poultrymen believe that, if a hen produces smaller eggs,
she consequently produces more eggs; and, conversely,
that if a hen produces larger eggs, she produces fewer
eggs. This matter has not been considered in the present
paper; it will be dealt with at a later time. The point
may again be stated, that the significant correlation exists
between numerical production and the ability on the part
of the hen to manifest an increase in egg-weight at those
seasons of the laying year when both production and egg-
weight attain their respective maxima. <A higher per-
centage of increase and absolute mean egg weight for the
entire year has not yet been attempted. Many points like
this.remain to be worked out and the author does not wish.
to present his results dogmatically, but only with the hope
that the problem will be attacked by other investigators.
It is not improbable that the results may be found to vary
with the breed of fowl, the date of hatching, the housing,
the feeding and with other factors.

With these points openly in mind, and only with the
purpose of stimulating further investigation and discus-
sion, the author presents the following brief summary as
expressing a biological fact which, if later proved to be of
general application, may take its place as a fundamental
law of production in the domestic fowl:

The innate egg-producing ability of a hen is manifested,
not only by the number of eggs laid within a year, or
within some shorter or longer period of time, but also by
the degree of increase or decrease in the mean weight of
her eggs, when this increase or decrease (calculated as a
percentage-increase or as a percentage-decrease) is meas-
ured at those periods of laying (the vernal and autumnal
maxima) characterized by the markedly increased pro-
duction of the flock; and on this basis, groups of hens
characterized by higher producing ability can be differen-
tiated as accurately as, and more easily than by any other
known means.

SOME HABITAT RESPONSES OF THE LARGE
WATER-STRIDER, GERRIS REMIGIS SAY

C. F. CURTIS RILEY

THe New YORK State COLLEGE OF FORESTRY AT
SYRACUSE UNIVERSITY, SYRACUSE, NEw YORK

CONTENTS

et

Introduction.
Physical Conditions of and Behavior in Brook Habitat, During
Severe Drought at White Heath.
x ES of Physical Daia:
Behavior in a Drying Pool.
x Hibernation and Æstivation.
4. Initial Responses in Dry Bed of Brook.
5. Behavior in Dry Bed of Brook.
III. Experiments in Connection with Brook Habitat at White Heath.
1. Methods.

=
P

2. Responses When Facing Bro
3. Responses When Parallel oe Biok
4. Responses When Te — from Brook.
5. Experiments with Bar
EY. fe aes of PENA in Connection with Brook Habitat
Syr

L Dt of Habitat.
2, Methods.
3. Responses When Facing rd from Pool.

Vy TERA of Observations at Whi
1. Initial Locomotor i

VI. Discussion of Tapina at White Heath.
1. Rôle Played by Vision.
2. Rôle Played by Moisture
VII. Discussion of Experiments at ‘Syracuse.
1. Rôle of Vision,
2. Role of Moisture,
VIII. Summary and —
IX. Acknowledgments.
X. Bibliograp

No. 628] HABITAT RESPONSES OF WATER-STRIDER 395

I. [NTRODUCTION

Tue large water-strider, Gerris remigis Say (Fig. 1),
is one of our most interesting and familiar aquatic bugs.
During the years 1911-1913 inclusive, I made a somewhat
intensive study of the responses of this species to the
physical conditions of its environment, in the vicinity of
Urbana, Ilinois. This study has been continued, inter-
mittently, up to the present time. The present publica-
tion forms only a part of the entire investigations.
Part of this paper treats of observations made near
Urbana, Illinois, and part treats of observations made
near Syracuse, New York.

Fic. 1, The large water-strider, Gerris remigis Say, natural size. (Folsom.)

In the summer of 1911, a severe drought occurred in the
vicinity of Urbana. In fact so extended was the period of
dry weather, that many streams in the region, that usually
were to be classed as permanent, became absolutely dry,
and others were reduced to a few isolated pools. Water-
striders, Gerris remigis, were trapped in many of these
pools. As the gerrids belonging to this species are mainly
apterous insects, they were unable to migrate by flight to
some other body of water. Because of this, serious results
might accrue to those individuals that were isolated in
such situations when the pools became dry. It was a
matter of interest to know what would become of these
hemipterons when the water entirely disappeared. There-
fore, a number of observations were recorded in connec-
tion with this subject.

396 ~ -THE AMERICAN NATURALIST (Von. LIII

II. Puysican Conprrions oF AND BEHAVIOR IN Brook
Hasrrat DURING Severe Drovenrt at Waite HEATH

1. Description of Physical Conditions.—Frequently
certain physical conditions were found to exist in brook
habitats in the early stages of droughts that had a direct
bearing upon the very existence of the water-striders,
Gerris remigis, but which will be mentioned here only
very briefly. I wish to refer particularly to a brook,
flowing partly through a forested region, near White
Heath. This brook is situated about eighteen miles
southwest of Urbana, and the physical conditions to be
considered are such as existed during a drought, in the
summer of 1911. During the earlier periods of this
drought, I often found, in the drying brook bed, small
pools of water not entirely isolated from each other. (In
Figs. 2 and 3 are shown drought stages in the bed of the
brook near White Heath that are very similar to those
under consideration. The only difference here, of im-
portance, is that the pools are in somewhat earlier drought
stages than are those to which I have referred.) Such
pools were connected by means of riffles not more than
6-12 inches wide (Figs. 2, 3). When food became scarce
in pools of this character, or when a scum, often due to
bacterial growth (Fig. 3), formed on their surfaces, I have
observed that the water-striders made their way from
one pool to another, by means of these small riffles of
water, until such connecting links disappeared and the
majority of the pools became dry. Eventually the gerrids
were concentrated on the surfaces of the few isolated
pools that remained. Sometimes, the bacterial growth,
which was of a gray color, anad the death of hundreds
of water-striders.

Such pools as have been mentioned persisted longer
in that part of the brook’s course that extended through
the wooded region. In such a region the pools were
larger, with a greater volume of water, than frequently
was the case in more exposed situations. These condi-
tions were due primarily to the protection afforded to the

No. 628] HABITAT RESPONSES OF WATER-STRIDER 397

Fig. 2. De tail of brook near White Heath, along margin of forested region,
showing EE during early stage of drought (July). Arrow indicates direc-
tion of current. a, pools on pies of which water-striders, Gerris remigis, live;
b, small vine se Suite two pools; water-striders pass from one pool to another
by means of such riffies. c, dry bed of brook exposed during early and late
stages 4 drought. pog

398 THE AMERICAN NATURALIST [ Vou. LIII

water by the surrounding trees. There, undoubtedly, is
less evaporation in a region of this character than is true
in those parts of the brook which are situated in regions
lacking both trees and shade and are thus exposed to a
high temperature and to the full effects of the summer
sun. Very similar facts were observed near Charleston,
Illinois, by Adams (1915, pp. 65, 66) in connection with a
small stream in a forested area. He makes the following
statement :

This small temporary stream in a ravine formed the southern bound-
ary of the area examined. . . . At the season of our examination it was
a series of small disconnected pools. . . . On the surface of the pools
were numerous specimens of a water-strider, Gerris remigis. The forest
cover is undoubtedly an important factor in the preservation of such
pools, as it controls the evaporating power of the air.

In this connection I wish to state that the water-striders
were found in far greater numbers, in the late spring,
summer, and early fall, on those portions of the stream
that flowed through the forested area. After several
years of observation, I have come to the conclusion that
the shade and lower temperature are the important faetors
in influencing the gerrids to remain in such situations.
De la Torre Bueno (1911, p. 246) has observed somewhat
similar facts, as is indicated in the quotation:

It [Gerris remigis] is to be found most frequently on running waters,
although it also frequents still, but to a less extent. . . . They congre-
gate in groups in shady, slow-moving parts of streams, at the tree roots
projecting from banks into the water, in the shadow of bridges, and in
general in almost any place where they have some shelter from the
burning rays of the summer sun.

This observer (1917, p. 201), again writing of Gerris
remigis, states that:

These beasties are common and familiar sights to the lover of the
quiet flowing waters running to the distant seas. In these haunts, in
some still little bay or moveless backwater, under a bridge, or in the
shadow of a tree, or in the cool recesses of an overhanging bank, you
may see remigis gathered in numbers, rowing silently about.
they rear large families and spend at ease the sultry eda

No. 628] HABITAT RESPONSES OF WATER-STRIDER 399

Fic. 3. Detail of brook near White Heath, along margin of forested region,
showing Serum pria a later stage of drought than indicated in Fig. 2
Arrow ind cu

(August). icates direction of current. a, n sur-
face, formed by piam growth, which frequently kills water-striders in rg
numbers; b, which water li urface
being free f > c all riffle connecting two Is; water pass

r èg scum; ¢, small rifi
from one p another by means of such rifles. d, dry bed of brook exposed
during early he tera stages of drought. (Original.)

400 THE AMERICAN NATURALIST [ Von. LIII

Attention has been directed to some of the physical con-
ditions that existed in a brook water-strider habitat in
the early stages of a drought. Certain features of such
a habitat in the later drought stages, will now be con-
sidered. So far as I am aware, there are no records in
the literature on aquatic Hemiptera, that describe the
responses of these gerrids, Gerris remigis, after the water
in their habitat actually disappears. The statements
given here are very condensed records taken, chiefly,
from my extensive field notes of the summer of 1911.

Late in the summer a trip was made to White Heath for
the purpose of examining a number of isolated pools, on
the surface of which water-striders were trapped, in the
bed of the brook near that place. Particular attention
was directed to one pool in which the water had evapo-
rated rapidly during the previous week. It was exam-
ined about’8:00 a.m. and at that time the dimensions of
the pool were approximately 12 X 5 X 4 inches. It was
evident that the pool would be completely dry before
night. Evaporation was taking place rapidly, as the
atmosphere had been very dry for several days. The
heat had been intense for a number of weeks, a tempera-
ture of 90°—100° F. not being uncommon.

- 2. Behavior in a Drying Pool.—There were twenty ger-
rids trapped on the surface of the pool, already mentioned.
The insects did not move about very much on the water-
film. They were very quiet, frequently remaining,
practically, motionless for several consecutive minutes.
There was no behavior on the part of the hemipterons
that indicated any attempt to escape from the unfavor-
able surroundings. The members of this species being
largely apterous forms and their optimum habitat being
permanent brooks and streams of moderate size, with a
current of medium velocity, it was, perhaps, not strange
that they evinced no responses that showed definitely
adapted behavior of a character suitable to cope with
such abnormal conditions and unusual habitat changes.
At 2:00 p.m. the pool was almost dry, and by 3:30 p.m.

No. 628] HABITAT RESPONSES OF WATER-STRIDER 401

there was little to indicate its location except a small
area of rapidly drying mud.

3. Hibernation and Aistivation.—A number of writers
—among others Uhler (1888, pp. 268, 275), MeCook (1907,
p. 265), and Kellogg (1908, p. 198)—on water-striders
have stated that during hibernation these gerrids bur-
rowed into the mud, under the banks of streams, at the
bottom of water under stones and roots of trees, and at the
bottom of the pool under roots or stones; and as some
investigators, Tower (1906, p. 245), for example, con-
sider hibernation and exstivation to be ‘‘fundamentally
one and the same process,’’ I was interested to observe
whether the gerrids would burrow into the mud, or into
some moist, sheltered spot under the banks of the brook,
and remain there in a quiescent condition, a state of
estivation, until the rains came. However, there were no
definite responses of such a character. I doubt whether
water-striders do æstivate in the true sense of that term
during periods of drought, although more evidence is
necessary before a definite statement can be made.

As will be noticed later, the gerrids do respond to con-
tact stimuli. Generally such a stimulus resulted in in-
hibiting locomotor activity and the insects remained
motionless with their bodies closely applied to some solid
object in the bed of the brook. This might prove to be a
piece of dry mud, a stick of driftwool, a stone, or a clump
of dead leaves. As a result of contact stimuli, sometimes
they would crawl under such objects as have been enu-
merated and remain there for a considerable length of
time. I have recorded observations which prove that they
stayed in such situations from a few seconds to thirty
minutes. The gerrids might remain in places of this
character for even a longer time, as I have noticed, on a
few occasions, that they were still there when I discon-
tinued my observations for the day. However, I was not
able to find them in the same situation the next day, even
after most carefully marking the place. On several other
occasions, I have made similar observations, but never

402 THE AMERICAN NATURALIST (Vou. LIII

was I able to discover the water-striders the following
day.

It is possible that there was a trifle of moisture in such
situations, and that it was this, acting as a stimulus, which
kept the gerrids there. But during these periods of
drought—accompanied, as they are, in the prairie regions
of Illinois with excessive temperature and glaring sun-
light—I have failed, positively, to find moisture in such
places, except under large objects, which were not present
in the bed of the brook. Frequently, however, I have
noticed that under these objects, in the brook bed, the tem-
perature was slightly lower and of course there also was
the shade. It is probable, also, that evaporation is re-
duced in such protected places. That certain arthropods
are sensitive to the evaporating power of air is known
from the work of Shelford (1913, pp. 85-102), who has
demonstrated that the yellow-margined millipede, Fon-
taria corrugate Wood, and the ground beetles, Ptero-
stichus adoxus and Pterostichus pennsylvanicus, respond
negatively to the increased evaporating power of air.
Therefore, lower temperature, shaded surroundings and
reduced evaporation may be the three factors that ac-
counted for the water-striders remaining in situations of
the character that have been mentioned. Of course if it
was proved that they stayed there day after day, this
would indicate the possibility of their tiding over a short
drought in such protected places. I have noticed on a
few oceasions, at least, that the gerrids have died, if
away from water, when exposed to glaring sunshine and
high temperature, during a drought. I also have ob-
served, when the water in my laboratory aquaria was
allowed to evaporate gradually, until it entirely dis-
appeared, that, sometimes in a few hours and at other
times in a few days, the gerrids died. This occurred, not
infrequently, when the temperature was only 85° F.

During hibernation, there is no question about water-
striders remaining quietly in one place for a long period
of time. But it must be recalled that the temperature is

No. 628] HABITAT RESPONSES OF WATER-STRIDER 408

low at such times, and that they frequently seek dry situa-
tions. The subject of hibernation will be discussed at
greater length in another paper. I may say, however,
that these water-striders do not hibernate in any of the
situations mentioned by the writers quoted, except under
the banks of streams, and then away from the water. The
positive thigmotactie responses of these gerrids have been
observed frequently during periods of hibernation. At
such times they formed tangled masses, which were due to
the water-striders crowding closely together. These facts
were recorded in my field notes as early as the winter of
1912-1913. It is interesting to recall that Essenberg
(1915; pp. 397, 400) has observed similar responses in the
ease of Gerris orba Stal.

4. Initial Responses in Dry Bed of oo — With refer-
ence to the water-striders in the dry bed of the brook at
White Heath, it appeared as if the gerrids might die right
where they were, for no movements occurred for approxi-
mately ten minutes after the water had disappeared.
Soon they began to move away. I can not state abso-
lutely what was the stimulus that caused the initial loco-
motor responses, although it probably was the total dis-
appearance of all moisture, as this was the only evident
change in the external conditions. Singly and in small
groups they jumped and walked in an ungainly fashion
from the site of the former pool. While it was evident
that the water-striders were less accustomed to locomo-
tion on the land than they were to locomotion on the
water, yet they made fairly good progress along the bed
of the stream.

The evidence that the initial locomotor responses of
the gerrids, away from the pool in which they had been
living, were due to the drying up of the water is further
supported by the following facts: Very frequently, I have
observed that when water-striders were removed from
the surface of a pool in a stream, or from an aquarium,
where they had been kept in captivity, and placed on the
ground or on some other solid surface that they at once

404 THE AMERICAN NATURALIST [ Von. LIII

became active. This was true even if the gerrids previ-
ously had been inactive. They would start to jump and
walk in a very agile, if ungainly, fashion, and made rapid
progress, especially if they were on a smooth and level
surface. It seemed that the transference from the water-
film to a solid surface—for all other conditions were as
before—was a sufficient stimulus to cause internal changes
in the insects, that set free energy in the form of loco-
motor activity. To induce this result there was but one
change in the conditions. In this connection I wish to
direct attention to a statement by Jennings (1906, p. 285) :
Often, of course, stimulation does rouse an organism to increased
activity. But even in this case the activity is due to the release of
internal energy. It may, therefore, continue long after the stimulation
which inaugurated the release has ceased to act. Such continuance thus
does not necessarily imply continued action of the stimulus. In many
eases the specific stimulus to action is only the change of conditions.

At this point, it may not be out of place to refer to cer-
tain observations which, possibly, may have a bearing on
the locomotor responses of the gerrids, after the drying
up of the pool of water. Abbott (1918, p. 234), in con-
nection with some responses of land isopods, Oniscus
asellus Linn., Porcellio. rathkei Brandt, and Porcellio
scaber Latreille, to humidity and evaporation makes the
following statement :

So far as observation shows, the effect on land isopods of exposure
to a dry atmosphere, including the first effect in desiccation experi-
ments, is an inerease of activity. This is a useful adaptation, provided
the activity carries them to other regions where moisture conditions ap-
proach more nearly the optimum. `

5. Behavior in Dry Bed of Brook. Sey ten yards
farther downstream there was another much larger pool.
Its dimensions were approximately 3 yds. X 2 yds. X 5 in.
I was interested to observe how many, indeed if any, of
the gerrids would reach it, and thus tide over the period
of drought, Of the twenty water-striders present, eight
of them went in the direction of the large pool, and
wandered in an aimless, awkward manner down the dry

No. 628] HABITAT RESPONSES OF WATER-STRIDER 405

channel of the brook. They first tried one path and then
another. These trials resulted in bringing some of the
gerrids to the. pool, but with reference to other water-
striders the trials were not so successful. I found that it
was a rather strenuous task to watch all the gerrids,
although some of them frequently would come to rest.
Without giving too many details, it may be stated that all
of the eight individuals, already mentioned, made their
way in a blundering fashion to the large pool of water.
The first one reached the pool in 5 minutes and 30 seconds,
and at once moved on to the surface-film. The time re-
quired by the others to reach the water varied to a con-
siderable extent, the average being 14 minutes and 30
seconds. The last individual, of the group of eight, to
reach the pool was fifteen minutes in making the journey.
Three gerrids out of the eight traversed the distance to
the pool with only a limited number of stops and devia-
tions from the straight path. However, this is not to
say that the journey was free from erratic movements,
on the part of these insects, for such was not the case.
The stops usually occurred when the gerrids came in
contact with some obstacle in their path. It was not at
all uncommon, when the insects moved away from their
stopping place, for the direction of locomotion to be
changed. The five other water-striders frequently came
to rest against stones and pieces of dry mud. Such
contact pauses were evidences of the thigmotactice pro-
pensities of the gerrids. They also wandered from side
to side of the stream bed, trying one direction for a
certain length of time, and then, through the influence
of some stimulus, trying another. However, they made
only a very few turns, directly away from the down-
stream pool during their entire journey. One water-
strider crawled under a piece of dried mud and remained
there for three minutes.

There were twelve other water-striders that have not
yet been accounted for. Four of these wandered up-
stream and from one bank of the brook to the other, first

406 THE AMERICAN NATURALIST [Vou. LIII

in one direction and then in another. Sometimes they
stopped with their bodies in contact with some obstacle
in their path. Frequently the direction was changed when
they renewed the journey. The gerrids did not turn down-
stream for any considerable distance during any of their
movements. After having observed them at intervals for
two hours, I decided that they were unlikely to reach the
water unless they eventually should wander to a pool
somewhere upstream. The gerrids had moved away from
the site of the former pool to such a distance that the
water-strider farthest upstream was twenty yards dis-
tant from the starting point. One individual crawled
under a piece of driftwood and had been in that situa-
_ tion for twenty-five minutes, when the observations were
discontinued. Another water-strider crawled among
some dead leaves and was there for twenty minutes
previous to the time when I left that place for observa-
tions elsewhere.

Of the eight gerrids that have not yet been mentioned, —
three of them stumbled toward the left bank and five of
them wandered toward the right bank of the brook. There
were various obstacles in their paths, such as small rocks,
pieces of dry mud, dead leaves and driftwood. The water-
striders frequently came to rest in close contact with these
objects—response to contact being a prominent feature in
the behavior of these gerrids—and at length two of those,
that had moved toward the left bank of the stream,
jumped into a large crack formed in the baked mud of the
bed of the brook. I observed these individuals from time
to time—for at least three hours—until I left for Urbana,
at about 8:00 p.m., and they were still in the same situa-
tion. I should state here that I searched for these two
gerrids the next day and was unable to find them; nor was
I able to discover the other four water-striders—to which
I already have referred—that had moved upstream. The
responses of one of the three gerrids, that in the first in-
stance had moved toward the left bank of the brook, have
not yet been described completely. After reaching the

No. 628] HABITAT RESPONSES OF WATER-STRIDER 407

bank, it began to jump downstream. This continued for a
distance of two yards, when it turned, facing the right
bank of the stream, and jumped in that direction for one
yard until it reached the middle of the brook channel.
The gerrid again turned to the left, this position directing
its head downstream, in which general direction it con-
tinued to travel, until after a number of stops, it even-
tually reached the large pool. In passing over its entire
route, this water-strider consumed fifteen minutes.

It has been stated that five gerrids wandered toward
the right bank of the brook, on first leaving the site of the
former pool. Three of these turned toward the middle of
the stream channel and jumped upstream for two yards,
turning to the left, they moved toward the bank of the
brook. On reaching it, first one and then the others
jumped downstream. After a number of pauses and
deviations to the right and left, two of them again turned
with their heads directed upstream and continued to jump
and walk in that general direction. After several erratic
movements and two stops, the third gerrid turned up-
stream. The insect jumped in this direction for three
yards and then stopped with its body in contact with a
small rock. It remained in that position for five min-
utes. In the meantime, the two other water-striders
had worked still farther upstream. On going back to
observe the gerrid that had been resting against the rock,
it was discovered that this insect had moved away and
was walking upstream. These three water-striders were
observed for 1 hour and 50 minutes and as there appeared
to be less and less possibility of their reaching the large
pool downstream, the observations were discontinued.
When the insects were noticed last, they had traveled
eighteen yards upstream away from the site of the former
pool, from which they originally came. One of them had
crawled under a piece of driftwood and had been there
for thirty minutes when I discontinued my observations.
These gerrids were sought carefully the following day,
but I was unable to find them.

408 THE AMERICAN NATURALIST [ Vou. LIII

The wanderings of two water-striders, out of the group
of five, have not yet been traced. Attention already has
been directed to the fact that, when they first left the site
of the former pool, they wandered toward the right bank
of the brook. After reaching this point, they turned
downstream. I lost sight of one of them, while observing
some of the other gerrids, and I was not able to find. it
again. The other water-strider proceeded downstream,
but frequently deviated from a straight path, and often
paused with its body in contact with various obstacles.
After wandering downstream for three yards, it made a
complete turn and its head thus pointed upstream. This
occurred as the insect was making a jump, it appearing to
lose control of the orientation of its body. The water-
strider moved upstream for two feet and then turned to
the left, walking in that direction for four feet. The
gerrid made a wide turn to the left, so that its head was
directed downstream. After much erratic wandering, it
eventually blundered upon the large pool of water, having
consumed forty minutes in completing the journey. It
was a task of considerable difficulty to keep all the gerrids
under observation, because, being almost the color of the
background, they were very inconspicuous.

I have records of a number of other instances, treating’
of my observations on the responses of water-striders
after the drying up of several other stream pools, at
different times and during different seasons. Some of
these pools were at distances of less than ten yards, while
others were eleven, twelve, and fourteen yards distant
from the original pool on which the water-striders lived.
It is true that, on the particular occasion to which atten-
tion has been directed and which has been described in
some detail, half of the gerrids entrapped on the surface
of the pool eseaped—when it became dry—to another and
larger body of water by making their way overland. But
in all the other instances that were recorded in my field
notes of similar phenomena, a very much smaller percent-
age of the entrapped gerrids was successful in reaching

No. 628] HABITAT RESPONSES OF WATER-STRIDER 409

other bodies of water, after their own pool became dry. I
will give a few examples: In one instance, only two of the
trapped water-striders out of eight reached the surface of
another pool, situated about eleven yards upstream. This
was not accomplished until the gerrids had consumed
much time in random movements, stops and deviations
from the straight path. At another time, fifteen gerrids
were trapped in a pool along the bed of a brook that had
become almost dry, owing to a prolonged drought. Five
water-striders out of the fifteen, or one third of the total
number, only, were able to find another pool of water
seven yards upstream from the point where they had been
isolated. These gerrids made many trials, errors, delays,
and turnings before they reached the surface-film. On
still another occasion forty water-striders were entrapped
on the surface of a stream pool. After the water had dis-
appeared entirely, the hemipterons left their haunts, and
stumbled along the dry bed of the stream. After much
wandering about, first in one direction and then in an-
other, six of the gerrids found a pool fourteen yards down-
stream from their former abode. The method of reaching
the water seemed to be due to a crude form of trial and
error.

I am inclined to believe that in periods of severe
drought, large numbers of apterous individuals die. If
the drought is a short one, and there is a certain minimum
of dampness under logs, stones, and clumps of dead
leaves, some of the water-striders may survive in such
situations. In regard to physical conditions of a some-
what comparable character, Kirkaldy (1899, pp. 109,
110) makes the following statement with reference to
Gerris lacustris Linné:

In small wayside pools or streamlets which dry up periodically, it is
obvious that the apterous inhabitants will die out unless their habitat
is sufficiently near to a larger stream or pond; a severe drought occur-
ring over a fairly large area, prolonged sufficiently to dry up all the
water within that area, would cause all the species represented in that
year by apterous individuals only to become extinct.

410 THE AMERICAN NATURALIST [ Vou. LIIL

After having observed the erratic, aimless character of
the responses of these insects, Gerris remigis, their ability
to reach some other distant body of water appeared doubt-
ful, and very largely dependent upon chance. If a larger
body of water should be very near to the haunts of the
gerrids, then there would be, through their random jour-
‘neys, a greater possibility of reaching such a situation.
In any case, their responses appear to be responses of
repeated trial and failure, until through many repeti-
tions, some gerrids eventually reach the water.

Ii. EXPERIMENTS IN CONNECTION -WITH Brook HABITAT
AT WHITE HEATH

1. Methods.——I now wish to record certain other ob-
servations, on the responses of. Gerris remigis, of a some-
what different character from those that already have
been discussed. Many times when out in the field collect-
ing Gerris remigis, some water-striders escaped while I
was picking them out of the net as it lay on the ground
near the margin of the stream. I noticed that these indi-
viduals generally found their way back on to the surface
of the water. Several times it occurred to me to under-
take a series of experiments with these water-striders for
the purpose of discovering their ability to return to their
own habitat after having been removed from it. The
_ plan was to place them on the ground at various distances
away from the stream and observe whether or not they re-
turned to the water, and if so to record the readiness with
which this was done. Up to the present, the experi-
mental work has not yet been carried out as extensively
as was desired, but the results that have been obtained
may be of interest. The experiments were undertaken
at the brook near White Heath, that has been mentioned
previously. The site selected was at a place where the
bank was flat, with its surface raised only a few inches
above that of the water.

2. Responses When Facing Brook.—Twenty gerrids
with their heads turned directly toward the stream, were

No. 628] HABITAT RESPONSES OF WATER-STRIDER 411

placed on the ground one yard away from the water.
All those that got back to the water succeeded in reach-
ing the surface-film in less than one minute. Similar
experiments were repeated for four successive times,
using different individuals. There was no experiment in
which more than two gerrids failed to reach the water.
Other experiments were carried out in which the water-
striders were placed two, three, and four yards away
from the stream. The majority of the gerrids found
their way back to the water. All those that were taken
two and three yards away from the stream were back
on the surface-film within 2 minutes and 30 seconds.
Gerrids that were placed on the ground four yards away
from the brook evinced a few more random movements
than was the case of those nearer to the water. A slightly
smaller percentage of the gerrids found the way back to
the stream from this distance. Those water-striders that
reached the surface of the brook did so within four
minutes from the time each experiment began.

3. Responses When Parallel with Brook.—A series of
experiments were performed identical to those already
described, except that the water-striders were placed on
a line with the long axis of their bodies parallel with the
bank of the stream. It was observed that the majority
of the gerrids were successful in reaching the water. In
the case of some individuals, a longer time elapsed than
was true in the first series of experiments. But most of
the water-striders either turned at once toward the brook,
or else they did so a few seconds after the first locomotor.
movements began.. In some instances the gerrids
jumped, for one yard or more, in the direction in which
their heads were directed, before turning toward the
brook. Some gerrids were placed on the ground four
yards distant from the water. Certain individuals of
these evinced some hesitancy in jumping directly toward
the brook and there were a number of random movements.

4. Responses When Facing Away from Brook.—Other
experiments, of a character similar to those that already

412 THE AMERICAN NATURALIST [ Von. LIL

have been described, were carried out, in which the heads
of the water-striders were directed away from the stream.
As before, a majority of the gerrids were successful in
reaching the water. The same distances were used as in
the former experiments, but in all cases there was less
promptness in moving toward the brook. It also was
evident that there were more random movements than
was true in the former experiments. Of the water-
striders placed four yards away from the stream, a
greater percentage, than in any of the other experiments,
failed to reach the water. 4

These hemipterons employed two methods of locomo-
tion on land, walking and jumping. The more rapid
progress was made by jumping. The individual jumps
averaged about one inch in length, although frequently
they were greater than this. Sometimes when jumping,
the gerrids made a complete turn, as if they were not
able, fully, to control the orientation of the body, and this
frequently resulted in changing their direction of prog-
ress. In this connection, I wish to direct attention to the
fact that Essenberg (1915, p. 399) has observed a some-
what similar response on the part of Gerris orba Stal and
makes the following statement:

It runs with a jerking motion, making from four to six jumps in
suecession and then making a short stop. Very often it turns a somer-
sault and continues running without interrupting its course until it
reaches a place of safety. There it lies quietly for from fifteen to
twenty-five minutes, then suddenly begins its race again.

Often during one of these jumping movements, the water-
striders struck against some obstacle and when this
occurred, their direction of progression was changed.
There was some tendency for the gerrids to continue to
move in the direction which they already had taken, with-
out regard to the position of the stream. In the three
series of experiments it is of interest to notice that the
water-striders reached the brook with a fair degree of
directness. It was only in the series of experiments, in
which the heads of the gerrids were turned away from

No. 628] HABITAT RESPONSES OF WATER-STRIDER 413

the water, that there was much evidence of lack of prompt-
ness in turning toward the brook and of random move-
ments.

5. Experiments with Barrier.—In the experiments that
have been described, I felt that the sense of sight was the
important factor in directing the water-striders to the
brook. However, there was always the possibility that
moisture, as a stimulus, might influence the responses of
the gerrids. Therefore, I planned a number of experi-
ments for the purpose of obtaining more definite infor-
mation on this subject. My idea was to arrange a sort
of barrier, along a limited section of the bank of the brook,
which was to be of such construction as to permit diffus-
ing moisture to pass through it readily, but on the other
hand, the barrier was to be so opaque that the reflection
from the water of either sunlight or diffuse daylight could
not be seen through it. The barrier first employed was
made of roughly interwoven leafy branches of trees. This.
was held in position by wooden stakes. It was four feet
high and extended along the bank of the stream for a
distance of approximately fifteen yards. There were
two reasons for employing a barrier of this character:
first, the readiness with which the materials could be
obtained; and second, the fact that its appearance was in
harmony with the general environment. However, it was
found to be practically impossible to make it sufficiently
tight so as to prevent reflections from the surface of the
water from passing through it.

A number of series of experiments were tried with
water-striders placed on the ground one yard and also
three yards away from the brook. While the experi-
ments were not sufficiently satisfactory, as a basis for
definite conclusions, it may be stated that the gerrids
displayed somewhat less promptness in reaching the
water than was the case when such a barrier was absent.
However, in general it was evident that the water-striders
jumped toward the brook with considerable expedition
and definiteness. If the barrier had proved to be light-

414 THE AMERICAN NATURALIST [ Von. LIII

tight, I should have been inclined to explain such a meas-
ure of promptness in moving to the water as due, perhaps,
to the effect of moisture from the stream.

Of course if the means of finding the brook was
mainly through vision, it was to be expected that the
gerrids would be a little less prompt in reaching the
water, because less reflection from the surface of the
brook could be seen than was true with the barrier re-
moved. Unfortunately, the movements of the leaves by
the wind permitted many bright reflections to pass
through the barrier. Lying flat on the ground, with my
eyes as close to it as practicable, in order to take a posi-
tion as nearly the same as possible to that of the water-
striders, I found that I was still able to see the bright
sunlight of the hot summer day, reflected from the sur- -
face of the water. I planned to use another form of
barrier and test this matter fully, when, because of the
character of the season, I was compelled to take up cer-
tain other observations, and I have not yet had the oppor-
tunity to supplement this work with additional experi-
ments. ;

(To be continued)

ONCHIDIUM AND THE QUESTION OF ADAPTIVE
COLORATION!

W. J. CROZIER
BERMUDA BIOLOGICAL STATION

AND

LESLIE B. AREY

NORTHWESTERN UNIVERSITY MEDICAL SCHOOL

I. As the result of his extensive study of the phenom-
ena of coloration in the tropical reef-fishes, Longley
(1917) comes to the conélusion that the colors of these
animals are ‘‘correlated with their habits in such a way
that their conspicuousness is thereby reduced. There is
no correlation of bright colors with special modes of de-
fence, and no evidence that bright-colored types enjoy
_ Immunity greater than that of their fellows.” The gen-
eral tendency of this, the most recent, and in many ways
the most thorough, observational study of the question of
adaptation in animal coloration that has yet appeared,
favors adherence to the view that as a whole the colora-
tion of animals is of a concealing, hence adaptively pro-
teetive kind. There are instances among invertebrates
which seem fundamentally inconsistent with this inter-
pretation, because the organisms in question are not
colored like their habitual background and because they
are for this, and for other independent reasons not incon-
spicuous. It is also a fact that in specific instances of
this sort special repugnatorial systems are frequently
encountered. Because of its importance for the theory
of adaptation, this matter is deserving of careful seru-
tiny. The natural history of the littoral pulmonate On-
chidium floridanum Dall affords a case in point.

1 Contributions from the Bermuda Biological Station for Resea
from the Anatomical Laboratory of the Northwestern University Medical
School.

415

416 THE AMERICAN NATURALIST [Vou. LIII

In his notes on the fauna of Bermuda, Heilprin (1889,
p. 187) described as new the form ‘‘Onchidium (Onchi-
diella) ‘trans-Atlanticum,’’ saying of its pigmentation:
‘‘ smoke color or dark olive,’’ in dorsal view. Some years
later Pilsbry (1900, p. 503), reporting on Verrill’s col-
lections at Bermuda, listed this species under the desig-
nation by which it is now known, Onchidium floridanum
Dall, adding, ‘‘Its dark olive green color agrees so well
with the stains on the rocks that it is very inconspicuous.
—(A.E.V.)”’

It is possible, though not certain, from these brief de-
scriptions, that Heilprin and Verrill had respectively col-
lected representatives of the two differently colored types
of Onchidium which occur at Bermuda. No constant
morphological differences have been detected upon which
to base a separation of these types, one of which is in
nature blue-black in general aspect, the other characteris-
tically of a light (‘‘smoky’’) olivaceous hue. Some in-
termediate shades do occur, but in the field a rather sharp
separation of the two groups is always possible. The
very dark form is much the more numerous. So far as
pigmentation is concerned, it is probable that the blue-
black types contain merely a heavier deposit of a sub-
epidermal pigment, which is chemically alike in both
kinds of Onchidia. Nevertheless, we are inclined to be-
lieve that the difference in pigmentation may be repre-
sentative of a real genetic difference between the two
types, for reasons which may briefly be stated as follows:
The two modes of colorization are in many places exem-
plified side by side; young and adults of maximal size are
found belonging to either group; the occurrence of the -
pale form is independent of the seasons, and is found
year after year in the same spots; in other places, where
blue-black individuals may be very numerous, no lightly
pigmented individuals have ever been encountered.
Therefore we are probably dealing, not with a temporary
fluctuation in color, determined by some metabolic cycle,
but with constitutionally diverse races of Onchidium;

?

No. 628] ONCHIDIUM AND ADAPTIVE COLORATION 417

whether they be ‘‘species’’ or not is immaterial. A simi-
lar phenomenon seems to occur in other genera of this
family, for v. Wissel (1898) notes that Plate (cf. Plate,
1893, 1894) collected two differently colored forms of
Onchidiella coquimbensis Plate; in this instance the color
difference between the two types, found in nature under
the same stones, was not so pronounced as we find it in
O. floridanum. In an Onchidiella which is very abundant
at Bermuda one of us has observed a corresponding sort
of color difference. The variation in question cannot be
in any way a sexual one, for reasons already given and
because the animals are hermaphroditic.

IL. Onchidium floridanum lives between tide levels. It
-is never seen when the intertidal shore-zone is covered by
the sea; during this period it inhabits deep crevices and
- cavities within the eroded rock. A dozen or more Onchi-
dia live together in this way. Their ‘‘nest’’ opens to the
exterior by means of a small opening, usually further ob-
structed by the growth of Modiolus, which is almost in-
variably quite inconspicuous, although the densely packed
small black ‘‘mussels’’ within and around the entrance
itself may form a very definite patch standing out clearly
amid the olive-brown alge covering the surface of the
rock. When the tide falls so far as to have left the level
of the Onchidium nest about an hour before, the occu-
pants of that nest emerge, successively and in a steady
stream, wander some little distance over the rock, feed,
and then, before the tide has risen Again, the members
of each group synchronously begin to creep back to the
respective nests from which they originated. The re-
markable character of this ‘‘homing’’ behavior we have
separately studied with considerable care, and have dis-
cussed in another place (Arey and Crozier, 1918). Here
we would emphasize simply the fact that O. floridanum
appears in the open, on the intertidal rocks, only during
the period of low water.?

2 The experimental observations from which some of these statements
are derived will be found described in a report, by Arey and Crozier, on the
general behavior of Onchidium, now in course of preparation.

418 THE AMERICAN NATURALIST [ Vou. LIII

The nature of the background provided on this rock
surface is by no means constant in different localities.
Commonly, as in the case of rocks in sheltered places,
completely submerged at high water, or on lee-shores
generally, the limestone is covered by a feltwork of
matted Enteromorpha and associated plants, sun-
bleached to an olivaceous yellow tint. This is the usual
condition within the sounds (Great Sound, Castle Har-
bor), and along the shores of bays protected from the
surf. In places not so sheltered the rock may be colored
by minute green alge, or by the growth of calcareous
green or reddish algæ, with which, as in the former case,
many serpulas are intermingled; barnacles (Tetraclita
porosa) also sometimes extend into the Onchidium zone,
but the prevailing hue, and the color of any particular
spot of small area, is largely determined by alge.

Onchidium also occurs, though more sparingly, on sur-
faces of a still different type. In some places the layer
of red earth (clay) underlying superficial æolian lime-
stone is exposed at the water’s edge, usually in a small
cove containing loose rocks left bare by the tide. This
earth provides a muddy layer of reddish brown, over
which Onchidia are occasionally found to be creeping.
From such tiny coves the direct rays of the sun are
largely excluded, so that the in-shore surfaces of the
stones are overgrown by alge, green or red-brown, un-
bleached; on these surfaces also Onchidium. may be
found. |

The zone inhabited by Onchidium is also that in which
great numbers of the common Modiolus occur. These
mussels, as already stated, almost invariably line the ap-
proaches to the Onchidium nest, and grow within the
entering passageway itself. Except where the mussels
have settled thickly in more or less horizontal fissures
between the wind-formed strata of the limestone, they
occur conspicuously in little groups, frequently not more
than 3-4 cm. in diameter, occupying every slight depres-
sion in the surface; this is usually the case in situations

No. 628] ONCHIDIUM AND ADAPTIVE COLORATION 419

where Onchidium abounds. A typical instance is that of
a low-lying islet such as ‘‘ Little Agar’s,’’ in Great Sound.
The leeward (eastern) side of this island, well protected
by the mass of the larger Agar’s Island, exposes at low
water a considerable extent of almost horizontal rock-
surfaces, whereon occur sparingly clumps of Modiolus, .
which are, however, much more abundant around the ver-
tical edges of these slabs. Within the clumps of Modio-
lus lie many Onchidium nests, from which the mollusks
at appropriate times creep out and wander over the flat,
yellowish, sunlit beach-rock. It is in situations of this
general type that Onchidium is most abundant; in less
freely illuminated places it oceurs more sparingy.

It is a remarkable fact that, in addition to frequenting
characteristically, though-not exclusively, habitats which
during most of the day receive the direct rays of the sun,
Onchidium comes out into the open only during daylight
hours, and never (in our experience) at night; the cir-
cumstance which makes this condition particularly
curious is that O. floridanum is at all times negatively
phototropic, and in a very precise manner, when tested
apart from its usual environment. No more striking in-
stance is known to us of apparent contradiction between
the findings of analytical experimentation regarding the
behavior of an animal and the actual life of the same
creature, for the snails when on the intertidal rocks move
toward or away from the horizontal rays of the sun with
perfect indifference. The cloudiness or brightness of the
day, however, has but an unimportant influence, if any,
on the numbers of Onchidia which appear upon the rocks.

That the wandering of Onchidium into the open is determined by imme-
diate physical circumstances in the environment is shown by two facts: No
rhythmic periods of migration persist when the snails are removed to the
laboratory; on days when strong winds and accompanying ocean yp ay
prevent the escape of the tidal water rer the partially enclosed Grea
Séund, so that there is no occurrence of ‘‘low water,’’ the idia remain

id
without much impairment; a small amount of air is sometimes trapped
within the rock cavities which they nr as nests and this may in some
cases help during prolonged submersi

420 THE AMERICAN NATURALIST [Von. LIII

III. The coloration of an Onchidium is the result of its
intensity of pigmentation, the texture of its mantle, and
the activity of its mantle glands. An adult measures, on
the average, 17 mm. long by 12 mm. broad by 6 mm. high
when resting undisturbed, being then oval in outline; but
it becomes more elongated during creeping, measuring
then about 24 by 8 mm., and<also less strongly arched,
being about 4 mm. high. When first observed creeping
out of its nest for an ‘‘airing,’’ it usually happens that
the Onchidium bears upon its dorsal surface a thin pel-
licle of slime completely investing the mantle. This
slime-layer is tough, somewhat thickened about the pe-
riphery, and can be peeled off intact. Frequently it is so
removed, mechanically, during the snail’s efforts to creep
out of its nest, for the passage-way is commonly so ob-
structed by the growth of Modiolus that the Onchidia
must slowly insinuate themselves through slit-like aper-
tures not more than 2 mm. wide. Occasionally Onchidia
are seen creeping about with their slime-coat but par-
tially removed.

The slime-pellicle is encrusted with the muddy fecal
matter which the Onchidia deposit. This calcareous
mud is swallowed when feeding on the alge over which
they creep in the open; it represents the ordinary silt of
the shore-line, and may in part be chemically abstracted
from the thin layer of sea-water covering the alge as the
tide recedes, owing to photosynthetic removal of CO»,
especially as accelerated by increasing temperature and
by isolation as the sea-weeds become more exposed.

When the slime-coat is removed, an ordinary O. flori-
danum seems at first sight jet black; the mantle is wet
and shiny; closer examination shows that the color is
really a very dark and smoky olive. The mantle-glands
continue their slime secretion, however, and by the action
of wind and sun the surface of the mollusk quickly as-
sumes a dry blue-black aspect, its numerous papille glis-
tening in the light; the appearance of the mantle, with
a slight ‘‘bloom’’ on it, then resembles that of the tiny

No. 628] ONCHIDIUM AND ADAPTIVE COLORATION 421

thysanurans so common on the beach sands and also on
the rocks where Onchidiwm lives.

Typical Onchidia of the lighter colored variety appear
at all times light olive or greenish yellow. The colora-
tion is not ‘‘solid,’’ for the dark-hued internal organs
usually show through as one or more dull blackish
patches. The edge of the mantle is quite pale.

It has not been possible to convert one variety into the
other by laboratory experimentation. The dark Onchidia
are still dark after six weeks complete starvation.

Onchidia of either type are found, frequently side by
side, in all the varieties of habitat which we have de-
scribed. To mention some specific localities, we may
list the following: Bailey’s Bay; Flatt’s Inlet; the east
side of the channel entrance to Hungry Bay, at its inner
end; an island at the entrance to Fairyland Creek; a
small cove on Burgess Point. These situations are
widely separated, and cover the different sorts of back-
grounds to which allusion has been made.

Not only in a general way, merely, but also in any
given place where they occur, the lightly pigmented On-
chidia fail to indicate by any peculiarity of behavior that
they differ essentially from the dark ones. That they
are easily overlooked is quite true; they are homochrom-
ically colored in practically all the instances that have
been observed. But it should be noted that in one such
place the homochromically colored individuals were
found in the proportion of but one to nearly 190 of the
dark black ones. This was on a greenish rock surface of
northern exposure where a color-changing form like the
isopod Ligia was pale-greenish, like the pale Onchidia.
In other places more or less isolated colonies of fifty or
more pale specimens were encountered. In some cases
the coloration was decidedly greenish, rather than oliva-
ceous yellow.

IV. The numerous dorsal papille of Onchidium were
said by Semper (1877, 1881) to indicate the location of
mantle-eyes; since, however, no sensory nerve-terminals

422 THE AMERICAN NATURALIST [ Vou. LIII

were at that time demonstrated within the ‘‘eyes,’’ some
writers followed Joyeux-Laffuie (1882) in the opinion
that the ‘‘dorsal eyes’’ of Semper were in reality mantle
glands, corresponding to those undoubted glands found
in the related genus Oncidiella (cf. v. Wissel, 1898). The
more complete analysis of the histology of the mantle in
Onchidium by Stantschinsky (1908), nevertheless, proved
that the ‘‘dorsal eyes’’ do possess nervous structures ap-
propriate for photoreception, the retinule being, in addi-
tion, of the inverted (‘‘vertebrate’’) type. O. floridanum
possesses, however, repugnatorial mantle glands as well,
which are not associated with the dorsal papille, these
last being relatively minute.

The glands are comparatively large, but are easily
overlooked in an undisturbed Onchidium. The periphery
of the mantle has a frilled appearance, numerous tiny
projecting papille giving it a serrated outline. If the
animal-be disturbed, the tissues surrounding each of the
glands become erected, and 14 stout conical papille, 7 on
either side and each with a terminal pore, become evident.

If the animal is suddenly removed from the rock by
grasping its dorsum, the glands all become turgid, point
upward and inward, and quickly discharge by the con-
traction of their muscular investments. The secretion is
a viscid fluid, milk-white, non-miscible with sea-water,
and decidedly acid in reaction. Discharge is more easily
elicited in air than when the animal is submerged. Under
water the secretion is expelled in long threads, but in air
the stream as it leaves the gland breaks up into a fine
~ spray, which may be shot to a distance of 12-15 em. With
vigorous stimulation, as when the back of the Onchidium
is pinched with forceps, the discharge appears from every
gland as a stream about 0.2 mm. in diameter, which is
often too stout to form fine droplets. In nature, how-
ever, the initial repugnatorial discharge is less copious;
it then takes the form of a fine spray, almost invisible,
which may be thrown for as much as 15 em., or about ten
times the length of the Onchidiwm.

No. 6289] ONCHIDIUM AND ADAPTIVE COLORATION 423

As discharged, the secretion is found to contain at
least three elements visible under the microscope: a clear
fluid, apparently water; clear droplets of oily appear-
ance; and a finely granular material. The relative pro-
portions of these three elements vary in different glands.
An individual tested in the field, without previous dis-
turbance, usually secretes a densely granular mass con-
taining a fair proportion of globules; in the activation, a
good deal of the secretion usually adheres to, or falls
back upon, the mantle of the snail itself—the dryness of
the mantle, and the fact that in nature violent discharges
of secretion seem seldom to be invoked are important for
the freedom of the mantle from being itself sprayed. On
the other hand, the fluids ejected from one or more
glands of starved Onchidia are almost always clear and
watery, sometimes with no trace of the granular or
droplet constituents. Variation in the composition of
the corresponding glands of Oncidiella was recognized
by v. Wissel (1898), who also discovered the secretion
droplets (globules) which previous workers had over-
looked.

When the secretion adheres to the mantle of an Onchi-
dium, it ean be seen that considerable stimulation of the
skin is thus produced. The secretion is quite acid to
litmus and to a number of other indicators. The acid is
found in the dense granular material, and does not occur,
save in relatively slight amount, when this material is
absent. The white masses may be caught on a clean slide
and studied under the microscope. The granular mate-
rial is insoluble in 90 per cent. alcohol, or in water. If
indicators are added to a drop of the secretion under a
coverslip, it can be seen that the acid slowly diffuses out
into the water from the granular mass.

The white substance, if received on the tongue, is found
to sting with considerable persistence, like wild mustard.
It does not taste sour. Plate, testing the secretion of the
mantle-glands of an Oncidiella, found ‘‘dass das Secret
dieser Driisen auf der Zunge ein deutlich wahrnehmbares

424 THE AMERICAN NATURALIST [ Von. LIII

Brennen hervorruft’’ (according to v. Wissel, 1898, p.
597).

Stimuli initiated in several different ways may serve to
elicit discharge of the repugnatorial glands, as: stroking
the mantle with a blunt point; pinching it with forceps;
applying solutions (5/8 M) of NaCl or (less effectively)
LiCl, or of methyl or ethyl aleohol (5 M). Touching or
pinching the periphery of the mantle itself does not lead
to discharge, but on the contrary leads to localized retrac-
tive puckering; when a gland is erected, preparatory to
discharge, stimulation of that gland causes it to collapse,
or ‘‘wilt,’’ a process similar to that following normal dis-
charge. If the dorsum be scratched until erection of the
glands occurs, and then NaCl solution be applied to one of
them, it is found, as also with mild faradization, that two
or even three successive discharges are possible; but the
later out-pourings of secretion are very meager as com-
pared with the first explosive ejection. About twenty-
four hours is required for the regeneration of a new
supply of repugnatorial substance.

Before the glands discharge, the edge of the mantle
curls upward, forming around the Onchidiwman encircling
saucer-edge bearing the now prominent glands; thus, de-
pending upon the intensity of excitation, the glands come
to point more and more dorsal, and in this position release
their contents. Not only so, but the individual gland
papille are themselves further directed in a general way
toward the irritated spot, so that if a slip of glass be held
horizontally 1 or 2 em. above an Onchidium the back of
which is at one point subjected to mild faradic excitation,
the contents of all the discharging glands are found to
impinge upon very nearly the same point; sometimes all
the discharges converge to an area of less than 3 X 3 mm.,
immediately above the spot irritated. The gland or
glands nearest the site of stimulation are the first dis-
charged, and the implication of more distant ones depends
on the degree of stimulation.

Semper (1881, p. 372 et seq.) considered that the dorsal

No. 628] ONCHIDIUM AND ADAPTIVE COLORATION 425

eyes of Onchidium were employed in sensing the approach
of animals which might attack it. He had in mind the
attacks of the Philippine Periopthalmus, a blenny which,
leaping along the intertidal beach-zone on its ventral
fins, feeds upon Onchidium and on arthropods. Semper
thought that the eyes of the Onchidia rendered them
sensitive to the shadow of an approaching fish, where-
upon a fluid spray was discharged from the skin glands,
driving off the fish. Whether Semper regarded the secre-
tion as repugnant to the fish in a chemical way, or whether
it was to act in a mechanical fashion merely, is not alto-
gether clear. Semper endeavored to make out a correla-
` tion, illustrative of evolutionary principles, between the
geographical distribution, on the one hand, of the species
of Onchidia possessing dorsal eyes, and, on the other, of
the genus Periopthalmus.

There are at Bermuda no Periopthalmus, nor any other
fishes of similar habits. Onchidium floridanum is, how-
ever, quite reactive to shading. The dorsal surface of
the mantle is the part sensitive to decrease of light inten-
sity. In a moving Onchidium the ensuing reaction in-
volves a momentary cessation of creeping, retraction of
the tentacles, and a depression of the periphery of the
mantle to the surface of the rock. The tentacles them-
selves, which in air are much further protruded than when
under water, are not photosensitive. An Onchidiwm
shaded when not creeping, but at rest or feeding, ex-
hibits the same type of response, but the mantle is more
decidedly humped. But, in our experience at least, no dis-
charge of the repugnatorial glands is induced by shading.
This does not mean that Semper’s observations were
not correct. In O. floridanum the type of excitation which
is preeminently and characteristically effective for gland
discharge is of a mechanical nature, namely, a rubbing or
pinching of the apex of the dorsum or a dislodging thrust
upon the ventral surface of the girdle. Although when
tested by faradic excitation it is found that gland dis-
charge may be elicited when even the sole of the foot is

426 THE AMERICAN NATURALIST [ Vou. LIII

irritated, or almost any other region (with difficulty in
the case of the tentacles), a greater intensity of stimulus
is necessary than when applied on the back of the animal.
Mechanical stimulation is in general more effective than
chemical. Of the chemical excitants tried, only NaCl,
LiCl, methyl and ethyl alcohols were successful, whereas
such substances as pure amyl alcohol, numerous acids,
alkalies and salts, in concentrations corresponding to
those used in the case of the successful stimulating agents,
as well as a series of alkaloids, were without effect when
small volumes (a half or one c.c.) were applied to the
mantle.

The designation of the mantle glands as repugnatorial -
is justified by observations in nature. Isopods (Ligia)
and grapsoid erabs (Sesarma, and in some places, Pachy-
grapsus) abound in certain spots where the Onchidia take
their airing. By prolonged watching, unequivocal in-
stances have been noted in which these arthropods have
come into direct contact with Onchidia, and in every case
the glandular discharge took place, followed immediately
by the retreat of the crab or isopod; it is not implied that
the crabs were endeavoring to eat the snails, but it seemed
rather that it was important for Onchidium to avoid being
accidentally pushed off the rock into the water; it does
not adhere to the rock with any great firmness, partic-
ularly under water, is easily dislodged, and if purposely
pushed over the edge of the stone on which it may be
creeping into the water, the animal finds it difficult if not
impossible to return to its own nest.

Tests upon a number of animals (e. g., Fundulus from
the land-locked brackish ponds at Bermuda) which could
not by any possibility have ever come into contact with
the Onchidium secretion, have invariably demonstrated
the powerfully repugnant character of gland contents.

Finally, it is inportant to note that Onchidia of what-
ever degree of pigmentation possess these repugnatorial
glands, and in an equivalent state of development. The
behavior of the differently pigmented individuals, with

No. 628]. ONCHIDIUM AND ADAPTIVE COLORATION 427

reference to the use of the glands, is identical throughout
the whole series.

V. Without entering upon a lengthy setting-forth of
the several alternative possibilities which might readily
be suggested in partial explanation of the foregoing series
of facts, we may state our belief that an attempt at such
. explanation in terms of color-adaptation is confronted by
serious, and indeed fatal, obstacles. From such a stand-
point we are required to consider only the period during
which the Onchidium is emerging from its nest and creep-
ing about in the open. In the case of a dark Onchidium it
is quite true that the animal is often very easily over-
looked while in process of creeping over the Modiolus at
the entrance of its nest; the color-match of the snail and
the mussel is a fairly exact one. But an Onchidiwm does
not resemble Modiolus, although of about the same size,
when creeping upon alge at the rate of 5 em. per minute
or faster. The paler kinds of Onchidiwm are, in almost
every instance noted, a very precise match for their back-
ground, and sometimes careful search is necessary in
order to detect them.

. The non-homochromiec pigmentation of the dark On-
chidia would, in view of the presence and behavior of the
poison glands, be of special interest in connection with
the idea of warning coloration. But this view is rendered
unintelligible in the face of the paler, concealingly-colored
types. Moreover, the Bermudian Oncidiellas to which we
have already alluded are very inconspicuous indeed, and
they also possess poison-glands; their mode of life is quite
different from that of Onchidium, since, although likewise
inhabiting the intertidal zone, they retreat during low
water to tiny holes,—dead serpula tubes and the like,—
coming forth again only when covered by the water.
These Oncidiellas are small, not more than 3 mm. long,
and both light and dark individuals are in the highest
degree difficult to detect when on their natural back-
ground; a delicate mottling of brownish pigment, in vari-
ous shades, has a pronounced concealing effect in these

428 THE AMERICAN NATURALIST [ Vou. LIII

places. Perhaps this accounts for the fact that Oncidiella
has not previously been reported from the Bermuda area.

Not only does the idea of ‘‘warning’’ coloration be-
come untenable for Onchidium, but the idea of adaptive
concealing pigmentation is likewise without adequate sup-
port. The dark Onchidia are not concealingly colored,
except during the transitory interval of their actual emerg-
ence from the nest, and even there they are easily seen, if
looked for. The slime-coating, which would undoubtedly
be of assistance in making them look like the background
(for the silt it contains is formed on rocks where the pale
bleached yellow alge grow, and in other places where
Onchidium lives), is almost invariably removed before
the snail begins its promenade in the open. The rela-
tively meager proportion of the Onchidia found pig-
mented in a truly homochromic manner is not explained
by the idea of adaptive coloration, although the variation
in the hues of differently tinted individuals would seem
to provide favorable ground for the operation of selection.
In this connection it might be suggested that Onchidiwm
floridanum represents a comparatively recent addition to
the fauna of Bermuda, and that the dark pigmentation
common to this species and its immediate relatives
throughout the world is even now in process of modifica-
tion; but the fact that in the neighboring genus Onci-
diella two more or less distinct general modes may be of
widespread occurrence forbids the placing of any special
emphasis upon this possibility, and even then, if correct,
it would have no final significance for the conception of
selective color-modification. If nutritive conditions, and
temperature or light, or all three, do operate in a manner
favorable to Onchidia of one or the other type in different
places, the habit of living together in colonies might lead
to a measurable degree of inbreeding, tending to prevent
the general distribution of one variety of pigmentation,
but at the same time a paler coloration, probably re-
cessive in genetic behavior, would to that extent achieve
a greater opportunity for perpetuation.

No. 628] ONCHIDIUM AND ADAPTIVE COLORATION 429

SUMMARY

VI. Onchidium floridanum Dall exhibits at Bermuda
two chief types of pigmentation, a pale type tending to
dull olive yellow, which tends to be concealingly colored,
and a much more abundant type of dark blue-black ap-
pearance.

No correlation can be established between the pig-
mentation of an Onchidiwm—which there is some reason
to consider the result of genetic factors primarily—and
the hue of the substratum over which the snail creeps in
. the open at low tide.

O. floridanum possesses repugnatorial martle-glands
of an effective type, secreting a granular emulsion of sub-
stances having a strongly acid reaction and producing on
moist surfaces of the human mouth a pronounced sting-
ing sensation. Touch and pressure stimuli on the dorsal
surface of the mantle are characteristically involveđ in
releasing the discharge of these glands, which shoot their
contents, in the form of a fine spray, to a distance about
ten times the length of the Onchidium and with conspicu-
ous accuracy of direction toward the source of excitation.

These facts are incompatible with the view that the col-
oration of Onchidium is determined or controlled by
selection in the direction of homochromicity or conceal-
ment. Nor can they be understood in terms of ‘‘warn-
ing’’ coloration.

REFERENCES

Arey, L. B., and Crozier, W. J.
1918. The ‘*Homing-Habits’’ of the sass apy Onchidium.
Proceed. Nat. Acad. Sci., Vol. 4, pp.
Heilprin, A.
1889. The Bermuda Isiands. Phila., vi + 231 pp., 17 pl.
Joyeux-Laffuie,
1882, Gidai et développemente de l’Oncidie. Oncidium celti-
cum Cur. Arch. Zool. expér., Tom. 10, pp. 225-388, pl. 14-22.
Longley, W. H.
1817. Studies upon the biological significance of animal coloration.
I. The colors and color changes of West Indian reef-fishes.
Jour. Exp. Zool., Vol. 23, pp. 533-600, 1 pl.

430 THE AMERICAN NATURALIST [ Von. LIII

agente A. ;
js air-breathing mollusks of the Bermudas. Trans. Conn.
Acad. Arts & Sci., Vol. 10, pp. 491-509, pl. 62.
Plate, L.
1893. Studien über opisthopneumone Lungenschnecken. II. Die On-
cidiiden, Zool. Jahrb., Abt. Anat., Bd. 7, pp. oe Taf.
7-1

1894. Mittheilungen über zoologischen ema an der chilenischen
Küste. Also Math. u. Naturw. Mitth. a. d. Sitzbr. f
pp. 59-67. Sitzb. Acad. Wiss. Berlin, ant 10, pp. 217-225.
gente K.
1877. Reisen im Archipel der P Theil 2, Bd. 3, Landmol-
lusken, Erzänzungsheft pp. 1—46, Taf. A-E. Viber Sehorgane
vom se us der Widens auf dem Riicken von

Schne
1881. haat! hit as affected by the Natural Conditions of pe
nee. Intern. Sci. Ser., Vol. 30, xvi + 472 pp. New Yor
Stantschinsky, W
1908. Uber Saar Bau der Riickenaugen und die Histologie der Riicken-
augen und die Histologie der Riickenregion der Oncidien.
Zeits. f. wiss. Zool., Bd. 90, pp. 137-180, Taf. 5-7.
Wissel, K. v
1898. nee zur Anatomie der Gattung Oncidiella. Zool, Jahrb.,
- 4

(Fauna Chilensis, Bd. 1), pp. 583-640, Taf.
34-36

THE CONTRIBUTION OF CARL FRIEDRICH VON
GARTNER TO THE HISTORY OF PLANT
HYBRIDIZATION

PROFESSOR HERBERT F. ROBERTS

KANSAS STATE AGRICULTURAL COLLEGE

In the beginning of the nineteenth century, the ques-
tion of the sexuality of plants was still undergoing a cer-
tain amount of disputation, despite Koelreuter’s investi-
gations; the history of which episode is sufficiently trace-
able through the writings of Schelver and Henschel. In-
order to assist in the settlement of the matter, the Royal
Prussian Academy of the Sciences at Berlin, made public
announcement in 1819 of an offer of a prize for the solu-
tion of the question—‘‘ Does hybrid fertilization exist in
the plant kingdom.’’ No response having been evoked,
the Academy extended the competition period, and
doubled the amount of the prize offered. On July 3, 1826,
the prize was awarded, although not in its entirety, since
the Academy did not consider the solution adequate, to
Dr, A. F. Wiegmann, of Braunschweig.

In January, 1830, the question was propounded anew
by the Dutch Academy of the Sciences at Haarlem, in the
following language:

What does experience teach regarding the production of new species
and varieties, through the artificial fertilization of flowers of the one
with the pollen of the others, and what economic and ornamental plants _
can be produced and multiplied in this way.

Since, by the termination of the contest period, January
1, 1834, no response had been received, the period was
extended to January 1, 1836. In October, 1835, Carl
Friedrich von Gärtner of Calw, son of a distinguished
botanist, Joseph Gärtner, formerly professor at the uni-
versities of Tiibingen and St. Petersburg, and who for a

431

432 THE AMERICAN NATURALIST [ Vou. LIII

considerable period of years had been conducting experi-
ments of his own in hybridization, became aware of the
offer of the Dutch Academy. Thus far, only brief reports
of his work had appeared at Tiibingen and Paris.

On account of the shortness of the time available,
Gärtner sent to the academy merely a preliminary report
of his experiments, accompanied by one hundred and fifty
mounted specimen sheets of his different plant hybrids,
which elicited a favorable response, and induced the acad-
emy to grant an extension of the time for sending in the
completed work to December 30, 1836. The interesting
resolution of the committee runs in part as follows:

That, in view of the number of new results, which could only have
been obtained through very manifold investigations over many years, the
service of the author be acknowledged.

The requirements of the committee having been com-
piled with, the award was formally conferred on May 20,
1837. The thesis appeared in Dutch translation, as a
document of 202 pages, in the Proceedings of the Academy
for 1838. In 1849, a revised and greatly enlarged edition,
‘‘the fruit of unbroken, zealous, almost twenty-five years’
work,’’ was published in German at Stuttgart.

The writer has found nowhere in current literature any
adequate presentation of this seldom read, and little
referred-to work, and yet it contains not only much inter-
esting information of a concrete nature, but a great deal
of speculative philosophical insight in dealing with prob-
lems in hybridization, that shows a scientific mind of
distinct value.

Sachs says of the writings of Gärtner (3), “The two works together,

are a brilliant termination of the period of doubt with respect to sexual-
ity in plants, which succeeded to the age of Koelreuter.” “ And thus it
was,” says Sachs further,—* in two small cities of Württemberg, that
the foundations of the sexual theory were laid, and the theory itself
perfected, so far as it could be, by experiment only, by three of the most
eminent of observers, Camerarius of Tübingen, Koelreuter and K. F.

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 433

Girtner of Calw, who contributed so largely to the empirical estab-
lishment of the theory, pan all that was done by others would seem of
small importance ” (p. 427).

s Focke says ((1) p. 437), “In numbers of experiments, he has
probably been surpassed by no other hybridizer.”

An idea of the sheer laborious work which Gärtner’s
operations involved, may be obtained from the mere state-
ment that he performed! close to ten thousand separate
experiments in crossing, involving nearly seven hundred
different species belonging to eighty different genera, and
from which some two hundred and fifty hybrid plants
were produced.

From such a large mass of detail as Gärtner’s memoir

involves, it is difficult to derive a concise series of state-
ments of the experimental results. In endeavoring to
group the phenomena of hybridization upon a scientific
basis, Gärtner undertook to classify hybrids into three
divisions, according to their external habit: (1) ‘‘inter-
mediate types,” (2) ‘‘commingled types,” and (3)
‘‘decided types,’’ although, as he says (p. 277):

There exists no exact delimitation among them, but they go variously
over into one another, so that it is not seldom very doubtful to which of
these forms, this or that hybrid should, with the greater right be
assigned.

In regard to the intermediate types, Gärtner follows
Koelreuter’s view

that, as in the fertilization of pure species, so àlso with hybrid breeding
in the case of simple hybrids, a complete balance occurred of both fertil-
izing materials, either in respect to mass or activity. In this assumption
he was still further strengthened through the similarity of types from
reciprocal crossing. He believed further, that in the later generations of
simple hybrids, and in the further grades of hybridization, where no
such regular process of hybridization occurs, the inclination of types
either toward the father or the mother, proceeded from the not quite
complete balance, or the slight overbalance of the one or the other
fertilization materials (p. 277).

Regarding the so-called ‘‘commingled’’ types, Gärtner
says as follows (p. 282):

434 THE AMERICAN NATURALIST [ Vou. LIII

The second kind of types, frequently occurring among hybrids, is
that with commingled parental characters, insofar as now this, now
that part of the hybrid, approaches more to the maternal or to the
paternal form, whereby, however, the characters of the parents in their
transference to the new organism, never go over pure, but in which the
parental characters always suffer a certain modification.

Under the third class of hybrid types, Gärtner considers
those—

Among which the resemblance of a hybrid to one of its parents, either
to the father or to the mother, is so marked and preponderating that
the agreement with the one or with the other is unquestioned, and
strikes one at once (p. 285).

Regarding the behavior of plant hybrids in the first
hybrid or F, generation, as compared with their behavior
in the second generation, Gärtner’s remarks are interest-
ing, although not based on numerical data.

Among plants, by far the greater number of the normal and regular

types of the hybrids from the first cross, as compared with the excep-
tions and varieties, testifies against the operation of . . . external influ-
ences, and proves rather the inner necessity of a regular formative
development according to law on the part of pure species in general as
well as of hybrids (p. 275).
_ Regarding this supposed formative force (Bildungs-
kraft), Gartner arrived at a conception of the inner
nature of the phenomenon occurring in hybridization,
which is scientifically interesting, although amounting to
a theory, and not to a conclusion directly derived from
experiment.

The foundation and the determination of the types of hybrids, might
therefore be discoverable, not so much in the mass and the relationship
of the germinal materials, as in a vital modification of the formative
force of the one or the other fertilization material (p. 270

As we now know from Mendel’s experiments, however,
it is this modification of the ‘‘sexual material’’ itself
which pre-determines the direction in which the forma-
tive force may run. In Girtner’s day, the abstract con-
ception of a relative potency of the one or the other
parent in hybridization, was a prevalent one, and this
statement above is a philosophical conception, which

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 4385

Mendel’s data made it possible to precipitate into a defi-
nite morphological theory. Gärtner went so far, how-
ever, as to attempt to measure this ‘‘potency’’ in a defi-
nite manner, by means of the number of viable seeds pro-
duced in reciprocal crosses.

That this relation of the factors of the fertilization forces in the case
of species of plants capable of hybridization, is definite and according to
-law, we assume from the fact that from such a hybrid combination come
more or fewer good seeds to be sure, according to the favorableness or
otherwise of the incidental circumstance, but that, however, in the case
of every such hybrid combination, it is never able to produce over a cer-
tain maximum of viable seeds, which maximum is peculiar to each com-
bination of that kind (p. 206).

Gartner goes so far as to say:

The inequality in the strength of sexual affinity (Wahlverwandt-
schaft), in the ease of the reciprocal combination of species, is a gen-
eral phenomenon occurring in plants, and will therefore lead in time to
the disclosure of the relations, and to a closer determination of the
value in respect to magnitude of the individual factors of sexual affinity.
This inequality establishes a scale of sexual affinity peculiar to each
species, which lies in the difference in the relationship of the strength of
the two factors. In this singular characteristic of plants, the peculiar
nature of the species is most plainly recognized,—much more, in fact
than in its external form (p. 200).

Those species which are able in crossing, to exert a
preponderating influence upon other species, Gartner
calls ‘‘generic types.”’

Just as such generice types in hybrid breeding, are able, as it were, to
gain the upper hand over the type of other species, so is the strength
and form of these species broken and overmastered by others (p. 290).

This manifestation of generic types, according to which one species
acts in a predominant manner over several other species in hybrid

reeding, is a further uncontradictable proof, that the relationship of
the forces, through which the union of two pure species takes place,
must be unlike, and that there can be no question there of any balance
of factors. To be sure in mixed hybrids, the relationship of the forma-
tive forces of the two sexual substrata appears to be tolerably alike;
however, in them likewise, in this or that part of the hybrid, now the
character of the one, now that of the other factor is more plainly ex-
pressed (p. 290).

Gärtner seems to see in this dominance of type in

436 THE AMERICAN NATURALIST [ Von. LIII

species, a way of evolution which leads to the establish-
ment of dominant family types.

The generic types appear to have their analogue in the natural family
types, and, since the origin of family types has occurred according to
certain laws, so will type-formation of hybrids, since it is not at
random, but is constant, follow the same laws, according to which plant
forms in general are formed and have developed (p. 291).

This brings Gärtner to a further statement regarding
the possibility of recreating ancestral types, which is
especially interesting as a pre-Darwinian view upon evo-
lution. Referring to species of Lobelia, Lychnis, Dian-
thus, ete., he says:

If these nearly related species had once come from a common ancestral
type, or had become separated from one another through the subsequent
development of the one or the other individual, then it appears to be
highly improbable that they would not again unite in their ancestral
form through reciprocal crossing, or prove themselves to be analogous
types in hybrid breeding (p. 163)

Gartner then arrives at one of the most striking con-
clusions in the older literature on the then so much
mooted species question, and which is the more interest-
ing because it takes the physiological rather than the
morphological point of view. He says:

The essential nature of a species, therefore, consists in the definite
relation of its sexual forces to other species, which relationship, taken
together with the specific form, is a characteristic, individual and con-
stant hie for every species, form and essence in this regard are one
(p.1

oe txoonantly returns to a philosophical contem-
plation of the nature of the fertilization process in hy-
bridization, characteristically as follows:

Not external resemblance in form and habit of species, but the har-
mony of the inner nature, gives the capacity for hybrid fertilization;
both are not always harmoniously bound together (p. 186).

For an actual hybrid combination, a certain harmony in both sexual
elements is, however, necessary, and precisely in this harmony lies the
eapacity for the union of two otherwise heterogeneous species (p. 110).

Only a harmony of the inner nature, on which the relationship of the
_ germinal materials rests, which indeed is ordinarily accustomed to be

associated with the external generic characters, but is not necessarily

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 487

bound up with concordance in external structure, determines the pres-
ence of the capacity for hybrid fertilization (p.

In the following discussion, more or less in the same
manner, Gärtner shows an intellectual freedom from the
fetish of morphological species, and clearly demonstrates
the possession of a physiological temper and attitude of
mind.

Our investigation concerning harmony of forms in families and gen-
era, have shown that we have to consider two different kinds of rela-
tionship among plants, an external and an internal: the former rests
upon conformity in habit, i. e., in growth, in the shape and form of the
leaves, and in the harmony of the flowers and the organs of pollination:
these, however, in their greater or lesser inclination to sexual combina-
tion of species in hybrid fertilization. The former might be identified
with the morphological, the latter closely with the physical relationship.
Now, since both are not infrequently met with in combination, for that
reason, our predecessors have not clearly distinguished both kinds of
relationship from one another, but have held both as identical, or rather,
have regarded the inner relationship as an immediate consequence of the
external, and assumed this as a law, so that the agreement of species in
habit, not only favored, not merely the existence, but also the strength
of the sexual attraction, and indeed likewise conditioned it (p. 166).

From the agreement of the external form and habit of species of
plants it may not therefore be inferred, that the sexual powers and rela-
tionships must also agree therewith, as experience indeed teaches, that
many plants, however congruent they otherwise are in form, and even
in organs of fertilization, nevertheless possess little or no inclination
to unite in hybrid fertilization. One of the best-known examples of
this is Pyrus and Malus, which, despite their near relationship in habit
and sex organs, from the testimony of other observers also, do not admit
of fertilization on the one side or the other (p. 167).

In the following passage, the morphological and the
physiological points of view are well contrasted:

The systematic genera are artificial syntheses, which are not united
according to absolute laws, but according to arbitrary external char-
acters, which indeed often harmonize with the inner nature of the
species, but likewise not seldom differ from this, if indeed it is not to be
questioned, that even these characters proceed from the inner organism,
and are determined through it (p. 139).

Gärtner, no more than any other ge in the
field of hybridization of his day, with the possible excep-

438 THE AMERICAN NATURALIST [Vou. LIII

tion of Sageret, had any conception of the idea of unit
characters operating as such, and capable of being ana-
lyzed separately. The then prevalent idea was one of
‘‘notency’’ and ‘‘pre-potency,’’ in the case of the hybrid
types that Gartner called ‘‘decided,’’ i. e., in which the
dominance of the one or the other parent was plainly evi-
dent. A species in a cross was supposed to function as a
whole as such. An idea of this older point of view is
obtained from the following:

Thus, just as there are species in a natural genus, which possess a
prepotent fertilizing power upon several other species of their genus, so
there are also species, which exert upon several others such a typical `
predominating effect, not to an equal extent, to be sure, but still of such
a nature that their operation in all combinations is to be recognized by

a character in common. Both of these forces, are, however of different
kinds, and follow different laws (p. 289).

Gärtner did not regard Sageret’s case of segregation
of characters to be the normal result of hybrid fertiliza-
tion. While it is true that Gartner recognized in a cer-
tain sense the fact that parental characters often behave
in a more or less unitary manner, he was led by the
nature of his mind, as well as by the results of his obser-
vations, to take a synthetic rather than an analytical view
of the hybrid organism.

The explanation of the origin and development of the forms of the
hybrids from the elements and characters of the parents, is as important
for plant physiology as for systematic botany (p. 25), and further,—

The laws of hybrid types orient themselves not toward the individual
organs of plants,—do not apply to a single part, e. g., stems, leaves, ete.,
but are applicable rather to the inner nature of species. The organs
which determine the types of hybrids, must therefore be investigated and
compared in their totality, and in their mutual interrelationship. For
the most part, the individuality of a hybrid expresses itself in its en-
tire habit, but in this respect, the flower above other parts of the plant,
is most frequently and plainly distinguished (p. 251).

However, Gartner’s most fundamental view upon the
question whether the plant as a whole, or its individual
characters considered as such, determine the nature of
the hybrid offspring, is expressed in the following clear
manner.

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 489

In the formation of simple hybrids, as in sexual propagation in gen-
eral, two factors are functional; this inequality of activity flowing out
of the specifie difference of species, expresses itself in the more marked
or the more feeble emergence of the individual paternal characters in
the different parts of the hybrid. Whether the species nature in its
entirety and its formative impetus, determines the direction and form
of the type, or whether also the individual parts of plants have a special
influence upon areon can only be determined through further
investigations (p. 257).

In the absence of what he deemed sufficient evidence to
the contrary, Gärtner conservatively adhered to the view
that the parent organism entered into the cross as a
whole, rather than as a congeries of character-units, be-
having in a manner separately, although often linked
together.

Regarding the recognized instability of hybrids, Gärt-
ner simply says, without distinguishing as to the gen-
eration:

Variability in the progeny of hybrids is a principal character of
hybridity (p. 518)

So far as any distinction as between the first and sec-
ond generation is concerned, Gärtner merely says:

The general laws of development of the parts of plants, hence appear
to undergo through hybridization no change perceptible to the ‘senses,
but all the developments and changes of the hybrid plant body appear
to follow the same laws as in pure species, the organs of reproduction,
and the material ground materials of the cross alone excepted.

The latter behaves differently in the second generation and in the
succeeding stages of hybrid fertilization, where, on account of the
` different nature of the two factors of the hybrid in the succeeding
zygoses, an altered, shifting, variable direction in type-formation enters
into the varieties thus originating (p. 572).

Concerning variability in hybrids of the second and
succeeding generations, he says:

Other hybrids, and in fact the most of them which are fertile, present
from the seeds of the second and further generations, different forms,
i. e., varieties, varying from the normal type, which in part are unlike
the original hybrid mother, or deviate from the same, now more, now
less (p. 422).

Perhaps the most definite allusion describing the con-

440 THE AMERICAN NATURALIST [ Von. LIII

dition in general terms, of what we term segregation in
the second generation, is the following:

Among many fertile hybrids, this change in the second and succeed-
ing generations, affects not only the flowers, but also the entire habit,
even to the exclusion of the flowers, whereby the majority of the indi-
viduals from a single cross ordinarily retain the form of the hybrid
mother, a few others have become more like the original mother parent,
and finally here and there an individual more nearly reverted to the
original father (p. 422)

Regarding the matter of unusual vigor in hybrids,
Gärtner remarks, giving examples, although again with-
out referring to any particular generation—

The marked inerease in the size of the flowers, is a phenomenon not
seldom occurring among hybrids (p. 295), and—

One of the most marked and general characters of plant hybrids, is
the luxuriance of all their parts, since among very many of them, an
exuberance of growth and development of roots, branches, leaves and
flowers manifests itself, which is not encountered among the parents,
even under careful cultivation (p. 526).

Gartner did not omit to apprehend the possible value
of this fact to agriculture, although, of course, he did not
recognize the first hybrid generation as a special phe-
nomenon.

Among the characters of hybrids Soio of recommendation for agri-
culture, their tendency toward luxuriance in stalks and leaves, and
their extraordinary capacity for tillering is related above. With respect
to the raising of forage, agriculture could, without doubt, make great
use of this characteristic (p. 634).

So far as genetics from the present technical stand-
point is concerned, Girtner’s data of course are not of
special interest, because his crossing was made upon
species as units, and not upon the character-unit basis,
and no records were made of the numbers of the different
types secured from his crosses. It is of interest to note,
however, that Gartner’s methods in his hybridization
operations partially anticipated the rigorous methods of

to-day, regarding the purity of parental types.

In order to judge with certainty concerning the nature of the types
which have arisen, and in order to obtain entirely reliable results, it is

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 441

above all necessary that one be in advance in complete certainty con-
cerning the species with which the experiments shall have been insti-
tuted, that they be specifically correctly determined, and that no doubt
prevail concerning their purity (p. 252).

Finally, Girtner’s investigations upon color inher-
itance, which cover thirty pages of text, while not of gen-
etic value from the modern standpoint, are interesting
and valuable as a summary of the then existing knowl-
edge on the subject. One observation upon intensifica-
tion of color deserves mention:

Red with red, not seldom gives a heightened brilliancy of color, as is
especially plainly shown in the flame-colored flowers of Lobelia cardi-
nalis, fulgens and splendens (p. 315).

One of the matters of genetic interest is the fact that
Gärtner experimented in the crossing of corn, with a
view to determining the matter of change of color in the
seeds due to crossing, as reported by Sageret. Unfor-
tunately for Girtner’s experiment, however, he crossed a
dwarf yellow corn without pericarp color, with corn hav-
ing colored pericarp (‘‘of red, gray and striped color’’),
instead of with colored endosperm. In consequence, of
course, the seeds borne the first year, ‘‘differed neither
in size nor in color in the least from the natural seeds of
Zea Mays nana of the earlier sowings’’ (p. 322).

The following year, however, instead of getting com-
plete color dominance, he obtained from one ear a ratio
of 224 with non-colored pericarp, to 64 with pericarp
colored. The other ear gave 104 seeds without pericarp
color to 39 colored. He carried the seeds through to the
next generation, but gives no numbers for them. Gärt-
ner also crossed Lychnis diurna with reddish or dark-
brown seeds, and Lychnis vespertina with ashy-gray
seeds, finding no change occurring as the result of cross-
ing, but obtaining what we should call dominance of ashy-
gray in the first hybrid generation. From Girtner’s ob-
servations, therefore, he felt justified in stating as alaw—

That the influence of the foreign pollen in hybrid fertilization, alters
nothing in the forms and external characters of the fruits and seeds

442 THE AMERICAN NATURALIST _ [Vow. LIII

peculiar to the mother plant, but produces in the embryo only, the
eapacity of bringing forth a mixed product from both concurrent
factors, through the germination and the further “development of the
new plant (p. 327

In an earlier paper of Gartner’s (2b), he cites Mauz’s
case of modifications in the character of different fruits
on a pear tree, through pollination from various varieties
of pears, whereby he was said to have obtained, ‘‘a great
number of fruits different in form and colors’’ (p. 138).

His interest. aroused by the phenomenon reported in
maize, he undertook a series of crossing experiments to
determine ‘‘ whether foreign pollen exercises or does not
exercise an immediate influence on the external character
of the fruits and seeds which are the result of these fer-
tilizations,’’ but with entirely negative results. No
change whatsoever was observed in the color or external
characteristics of the fruits arising from crossing.

The iffluence of the foreign pollen does not then change anything in
the external forms peculiar to the mother plant, or in the external
qualities of the fruits, the seeds and even the embryo. This influence
only gives to the latter the faculty of producing, through germination
and through the ulterior development of the new plant, an intimate
combination of the form of the members of the two species which have
united in its production (p. 139).

One of the most interesting matters, of course, is that
which concerns the alteration in the character of hybrid
seeds or fruits due to the immediate effect of foreign
pollen. Gärtner reviews in detail the previous work of
Knight, Goss and Seton with peas. In 1829, he started
a selfed and a crossed series of peas, using four varieties
(pp. 81-85).

1. Paris Wax (yellow seeds).

2. Dwarf Creeping (white flowers, yellow seeds).

3. Sugar peas (red flowers, wrinkled greenish-yellow seeds).

4. Early Green Brockel (white flowers, green seeds).

The results as to the immediate effect of the cross on
the seeds were as follows:

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 448

Parents. Hybrid Seeds.
Paris Wax (yellow) X Sugar Peas (greenish yellow)......yellow.
Paris Wax (yellow) X Early Green Brockel (green)....... greenish- ipt

Sugar Pea (greenish yellow) X Dwarf Creeping (yellow). .dirty-yel

Sugar Pea (greenish yellow) X Early Green Brockel (green).n oe

Dwarf Creeping (yellow) X Early Green Brockel (green). pran
irty-yello

Early Green Brockel (green) X Sugar Pea (greenish yellow) . yellow.
Early Green Brockel (green) X Dwarf Creeping (yellow). .. yellow.

As the above results show, the same dominance of yel-
low over green in the hybrid seed appears as in the ex-
perience of Knight, Goss and Seton.

Respecting identical results obtained as the result of
reciprocal crosses, Gärtner makes the following unquali-
fied statement:

The most important and the most interesting phenomenon in the
crossing of plants in hybrid breeding, is the complete similarity of the
wo products; since the seeds produced from the one as from the other
fertilization, give rise to plants of the most complete similarity, so that
their different origin and derivation, upon the most careful investiga-
tion of both kinds of hybrids, does not show the least difference in
respect to their form and type; and even the most practiced specialist
with a hybrid species, is not in position to distinguish the origin of the
hybrid with respect to the sex of the parents (p. 223).

Gartner’s work is not only noteworthy for its remark-
able extent with respect to the number of species experi-
mented upon, but with regard to the care which he exer-
cised in his operations, he says:

For complete assurance of the purity and reliability of the products
of hybrid breeding, and for testing the conclusions derived therefrom,
we have repeated most of the experiments, especially the doubtful cases,
not once only, but several times, and put them to the test through cross-
ing of the same species, using different individuals of that species, for
even with the most scrupulous foresight and precision, individual rare
instances have still occurred in these tedious and wearisome investiga-
tions, where the suspicion had made itself felt, of a mistake or error
having crept in, either in pollination or emasculation, since such results
stood in direct contradiction to the usual experiences, and on a repeti-
tion of the experiments, made itself incontrovertibly evident as an error.
We believed it possible to attain no higher degree of certainty in this
branch of natural science, and to be able to bring the conclusions derived
therefrom to no higher proof, than through the precise coincidence of

444 THE AMERICAN NATURALIST [ Vou. LII

the forms of the products, by repetition under the same conditions with
‘the same species, but with different individuals, and at different times
(p. 675)

Again, in another place, he makes substantially the
aes statement with = to testing what he refers to
as ‘‘selective affinity.’’

In order to gradually get as close as possible to the true selective
affinity relation among the species of plants, it is necessary not only that
a greater number of experiments be instituted with the same species of
plants, but also that the same experiments be repeated with different
individuals and at different times, because as well in the female organs
of a plant, as in the pollen of another species, a basis for different
results may be concealed (p. 214

Girtner was not behind in realizing the practical utility
of hybridization in agriculture and horticulture. The
following somewhat extended extract shows his keen
sense of interest to the possibilities latent therein, al-
though his own scientific efforts did not lead him into
economic experiments.

The heightened fruiting capacity of hybrids and variety crosses, de-
serves the most marked attention in respect to orehard, vineyard, and
the whole of garden culture. The striking fertility of several orchard
and vineyard varieties may find its explanation herein. It is, to be
sure, to be surmised that this capacity does not reside in an equal degree
in all variety crosses, and that this character would incline toward the
peculiarity of the species; nevertheless it is to be expected with tolerable
certainty, that with many valuable orchard and vineyard varieties, an
inereased yield might be able to be attained through crossing with other
varieties. Improved sorts with weak or weakened vegetative power,
united with other species of more vigorous growth, would promise an
improved product with a longer life duration and a stronger structure
of the plant body. As already many admirable stone, pome and vine-
yard varieties have been raised from seeds which had originated through
chance crossing, so, through intentional artificial crossing of varieties,
still many other sorts might be very easily produced. But to get definite
results, and to be able to determine the outcome exactly, for the advan-
tage of science as a whole, fertilization should not be committed to
mere chance, but an exact and scrupulous procedure must be observed.
with careful records of the varieties combine

Of still more extended utility is hybrid bebddiie for wsthetie botany;
for the latter, artificial fertilization opens a wide field for activity, en-
joyment and achievement. For the fancier of ornamental plants, the

No. 628] VON GARTNER AND PLANT HYBRIDIZATION 445

ease with which many hybrids are able to be produced, is an inex-
haustible source of satisfaction and profit. He recognizes that he is in
possession of materials with which he can busy himself, and he delib-
erates over the way and manner he can best and most profitably combine
them; in that, he gives attention to the characters wherein each species
characterizes itself, whether in the splendor of the colors of the flowers,
the fineness of their delineation; fragrance, growth, form, quantity of
flowers; whether endurance of the severity of our climate in this or that
combination is to be taken into special consideration ;—he will attempt

the hybrids, and he will finally be surprised at getting a plant which
had never before existed in nature (pp. 638

This concludes the matter of general interest in Gärt-
ner’s memoir. The writer believes that it should be
carefully read by every plant breeder, not only for the
details of practical and historical value therein contained,
but because of the philosophical spirit underlying Gärt-
ner’s scientific attitude upon the nature of the hybrid
organism.

BIBLIOGRAPHY
J. Focke, Wilhelm Olber

Die Sana. ein Beitrag zur Biologie der Gewiichse. Berlin

(1861
2. Gärtner, Carl Friedrich von.

a. Versuche und Beobachtungen über die Befruchtungsorgane der vol-
kommeneren Gewichse, und über die natürliche und künstliche
Befruchtung durch San eigenen Pollen.’’ Naturwissenschaft-
liche Abhandlungen. Tübingen, 1:

b. 1827, Notice sur des expériences concernant la fécondation de quel-
ques végétaux. Annales des sciences naturelles, 10: 113-148.
(Translation of the preceding.)

c. 1838. Over de Voorteling van Baastard-Planten. Eene Bijtrage tot
de Kennis van de Bevruchting der Gewassen. Haarlem.

d. 1844. Beiträge zur Kenntniss der Befruchtung. uttgart.

e. 1848. Versuche und Beob E über die Bastarderzeugung im

f. 1849. Methode der ERPE Bastardbefruchtung der Gewächse,
und Namensverzeichniss der Pflanzen, mit welchen Versuche
naar ellt wurden. Stuttgart.
8. Sachs, Julius
History of ‘iaiy (1530-1860). Trans. by Garnsey and Balfour.
Oxford, 1906.

ON THE USE OF THE SUCKING-FISH FOR
CATCHING FISH AND TURTLES: STUDIES
IN ECHENEIS OR REMORA, II.

DR. E. W. GUDGER

AMERICAN MUSEUM or NATURAL History, New York Ciry

II

In 1507, there was published at Venice by Franconzio
another collection of travels, entitled ‘‘Paesi Nouamente
Retrouati Et Nous Mondo da Alberico Vesputio Floren-
tino Intitulato’’ [Countries Newly Found and the New
World of Albericus Vesputius Called the Florentine].
Chapters LXXXIIII to CXIII faithfully reproduce the
Libretto of 1504, are in fact a second edition of the Li-
bretto, and need not detain us.1*.

All this however simply pushes the question back one
step further and it now becomes ‘‘What or who is the
source of Peter Martyr’s information?’’ The answer is
that these sources are identical with those for the ‘‘ Life
of Christopher Columbus’’ by his son Ferdinand, for Las
Casas’s history of the West Indies, and for Bernaldez’s
“Reyes Catalicos.’’ In addition Dr. Eastman has skil-
fully worked out certain internal evidence which points
directly to the one person who gave to Peter Martyr the
data incorporated in Chapter XV of the Libretto.

Let us first of all consider the account which Ferdi-
nand gives of the fisherman-fish incident, which seems to

‘14 The ‘*Paesi’’ was also reprinted at Milan in 1508 by Arcangelo
Madrignano, and later at both Basle and Paris by Simon Gryneo. Martyr
says that this plagiarism was the work of Alvico de Cadamosto and de-
nounces him in Decade II, book 7. Martyr’s ‘‘First Decade’’ was pub-
lished in 1511, and the first edition of the ‘‘Decades’’ appeared in 1516
under the editorship of the author’s friend, or tink de Nebrija. The pub-
lisher seems to have been Alcala de Henor

446

No. 628] STUDIES IN ECHENEIS OR REMORA 447

have occurred on or about May 19, 1494. This is found
in his ‘‘ Historie’’ as follows:

The nearer they sailed to Cuba, the higher and pleasanter the little

islands appeared which were all over that sea, and it being a matter of

diffieult d to no purpose to give every one of them a name, the
Admiral called them all in general Jardin de la Reina, the Queen’s
arden. . . . In these islands they saw crows and cranes like those of

Spain, and sea-crows [gulls], and infinite numbers of little birds that
sung sweetly, and the air was sweet as if they had been among roses,
and the finest perfumes in the world; yet the danger was very great,
there being such abundance of channels, that much time was spent in
finding the way out.

In one of these channels they spy’d a canoe of Indian AE who

very quietly, without the least concern, awaited the boat which was
making towards them, and being come near, made a sign to them in it
to attend till they had done fishing.
. Their manner of fishing was so strange and new to our men, that
they were willing to comply with them. It was thus: they had ty’d
some small fishes they call Reverso by the tail, which run themselves
against other fish, and with a certain roughness they have from the head
to the middle of the back they stick fast to the next fish they meet; and
when the Indians perceive it, drawing their line they hand them both
in together. And it was a tortoise our men saw so taken by those fisher-
men, that fish (the Reverso) clinging about the neck of it, where they
generally fasten, being by that means safe from the other fish ee
them; and we have seen them fasten upon vast sharks.

When the Indians in the canoe had taken their tortoise, and two other
fishes they had before, they presently came very friendly to the boat,
to know what our men would have, and by their directions went along
aboard the ships, where the Admiral treated them very courteously. .

The close similarity between the accounts of Ferdi-
nand and Martyr will already have occurred to the
reader. Before going further it is but fair to say that
Ferdinand’s original copy is not known, the printed text
being from Ulloa’s Italian translation. However, Fer-
dinand was the heir of the admiral, and, since all his
father’s papers which were preserved seem to have
fallen into his hands, may be considered as his father’s
literary executor. Winsor says (pages 9-10) that ‘‘Fer-
dinand, or the writer of the ‘Historie,’ ... it seems
clear, had Columbus’s journal before him.’ Columbus

448 THE AMERICAN NATURALIST [ Vor. LIII

kept a journal of his second voyage until he was stricken
down by sickness, and this is attested to by both the His-
torie and by Las Casas.

Ferdinand himself on this point says (Churchill’s
Voyages, II, p. 560) that, after the fishing scene above
described, the Admiral held on his course though worn
out with fatigue, neither having had his clothes off nor
lain in a bed since leaving Spain ‘‘till the 19th of May.
(1494) when he writ this,’’ i.e., the account of the fishing
scene in Queen’s Garden. Windsor states (page 39) that
the ‘‘ Historie’’ was up to 1871 believed to be a biography
of Columbus by his son Ferdinand, and that though
doubted by some, is still firmly held to by many authori-
ties. With the above conclusions, Dr. Eastman and I,
after a careful study of all the available data, found our-
selves in full accord.

Las Casas, the great apostle to the Indians, left at his
death a manuscript history of the West Indies. This
had been long in the writing, from 1527 possibly, or more
positively from 1552, to 1561 (Las Casas died 1566), but
was longer in getting published (1875). However, in its
manuscript form, it was available from the time of his
death for all later historians.

Las Casas’s account of the fishing scene need not de-
tain us here since it is essentially like that in the ‘‘ His-
torie” by Ferdinand, and like that in the ‘‘Libretto’’ of
1504 and the ‘‘Deecades’’ of 1511. Much more important
is the query as to the source of Las Casas’s data. Win-
sor (pp. 39 and 47) quotes Harrisse that he thinks that
both Ferdinand, or the author of the ‘‘Historie,’’ and
Las Casas had access to common documents or may be a.
manuscript prototype of their writings. And later (p.
56) Winsor speaks of ‘‘the journal of Columbus as pre-
served by Las Casas.’’

One further source of information needs to be set
forth, and then after a brief consideration of Peter Mar-
tyr’s sources, this part of our study will be finished.
About the middle of April, 1915, Dr. Eastman got word

No. 628] STUDIES IN ECHENEIS OR REMORA 449

of a manuscript copy in the Harvard Library of a manu-
script document in the Royal Library at Madrid of
Columbus’s time written by a personal friend of the
great navigator and narrating the events of the second
voyage. A few days later he wrote me as follows:

Everything run down thus far is overshadowed in importance by the
new find, ante 1500, which I take to be the ipsissima verba of Columbus
himself. The MS. ... now in the Harvard College Library formerly
belonged to Mr. Prescott, who had it transcribed from a MS. work in the
Royal Library of Madrid. A part of it was translated, rather poorly,
in the Massachusetts Historical Collection before 1850, and some years
later the Madrid MS. was printed (1856 at Seville and 1870 at Madrid).
Irving and Humboldt both consulted the original MS. or copies of it,
and historians agree that the author, Bernaldez, an Archbishop of
Andalusia, not only entertained Columbus at his house on his return
from his second voyage, but received the journals and other papers then
in Columbus’s possession. Prescott makes this statement and it is re-
peated by others. Now Bernaldez, in his work written before 1500
embodies practically all of Dr. Chanea’s'® letter, and hence we may
suppose that what he takes from Columbus’s papers and journals was
copied nearly verbatim. I regard this as one of the most important
authentic sources for the second voyage . . . coming as it does nearest
to the fountain head.

Let us now consider Bernaldez’s account, which as just
shown seems to be a transcription of Columbus’s own
words.

The Admiral set sail [from Jamaica] with his three caravels, and
sailed 24 leagues towards the west, as far as the gulf Buen Tiemps. . .
On Whitsunday, 1494, they stopped at a place which was uimh
but not from the inclemency of the sky, or the barrenness of the soil, —
` in the midst of a large grove of pam-trees, which seemed to reach from
the sea-shore to the very heavens. . . . Here they all rested themselves
upon the grass about these fountains, enjoying the charming fragrance
of the flowers, and the melody of the song of birds, so many and so
sweet, and’ the shade of the palm trees, so tall and so beautiful, that
the whole was a wonder. . . . As the number of islands in this region
was so great that he could not give to each a separate name, the Admiral
called them all by the common name of the Queen’s Garden

On the day following, the Admiral being very desirous fe fall in with

15 Dr. Chanca was a physician who accompanied Columbus on his second

voyage, and who wrote back a long letter describing various natural objects
in the New World, but saying nothing of the Remora.

450 THE AMERICAN NATURALIST [ Vou. LIII

some natives with whom he might parley, there came a canoe to hunt
for fish :—for they call it hunting, and they hunt for one fish with others
of a particular kind. They have certain fishes which they hold by a line
fastened to their tails, and which are like conger-eels in shape, and have
a large mouth [i. e., head] completely covered with suckers, like the
octopus. They are very fierce, like our ferrets, and when they are
thrown into the water they fly to fasten themselves upon whatsoever
fish they may ga eg sooner die than let go their hold till they are
drawn out of the w

The hunting fish j is very light, and as soon as he has taken hold, the
Indians draw him by the long cord attached to his body, and in this
manner they take a fish each time on drawing both to the surface of
the water.

As these hunters were at a distance from the caravel, the Admiral
sent his boats to them with armed men, contriving it so that they should
not escape to the land. As the boats came up to them, these hunters
called out to the men in mildest manner and as uneconcernedly as if
they had known them all their ives, to hold off, because one of the
fishes had fastened upon the under side of a large turtle and they must
wait till they got it into the canoe. This our men did, and afterwards
they took the canoe, and those in it, together with four turtles each of
which was three feet in length, and brought them to the ships of the
Admiral; and there they gave some account of these islands, and of their
cacique who was close at hand, and had sent them to hunt. They asked
the Admiral to go on shore, and they would make for him a great feast
and would give him all of the four turtles they had caught.

Now for a short consideration of Peter Martyr’s
sources, which seem to be in common with those of Fer-
dinand Columbus, Las Casas and Bernaldez, if we may
judge by the marked similarity of the accounts. There
can be no doubt that Martyr, who during all the years of
Columbus’s voyages, was an attendant at the Spanish
court, knew Columbus personally and held converse with
him about his voyages and the wonders seen thereon.
Winsor says (page 34) that ‘‘ Peter Martyr knew Colum-
bus,’’ and adds that ‘‘Las Casas tells us how Peter Mar-
tyr got his accounts of the first discoveries directly from
the lips of Columbus himself and from those who accom-
panied him.” And on the next page (35) we read ‘‘ Mar-
tyr ... composed a special treatise on the discoveries in
the New World . . . under the title ‘De Orbe Novo’ .

No. 628] STUDIES IN ECHENEIS OR REMORA 451

(which) occupied his attention ... till the day of his
death. For the earlier years he had... not a little help
from Columbus himself.’’

Let us now see what Thacher, the latest and most pro-
found of the biographers of Columbus, has to say as to
Peter Martyr’s sources, and we have done with this part
of this paper. On p. 215 of volume II (1903), he says:
“The Admiral and some of his followers wrote to Peter
Martyr, and Peter Martyr thereupon wrote [a series of
letters] to an Italian Duke and to a few Cardinals.’’ On
p. 218 ‘‘... Peter Martyr, who not only had access to all
public documents, but who himself corresponded with
Columbus.” On p. 440 Thacher referring to Peter Mar-
tyr speaks again of ‘‘. .. Personal correspondence with
the Admiral.’’

Confirmatory of all the preceding it may be noted that
the Spanish Jesuit, Nieremberg, professor of physiology
in the Royal Academy of Madrid, in writing of the Re-
versus, quotes Christopher Columbus. It seems not un-
likely that he had in his day (his book was published in
1635) access to some of the Columbus manuscripts, may
be to the journal of the second voyage. And earlier than
Nieremberg, Gesner (1558) on page 483 refers to ‘‘ Christ.
Colibus’’ as his authority for the story of the hunting
fish. Furthermore Humboldt (1826) quotes Columbus
on the activities of the Reves.

From a consideration of all this testimony, no other
conclusion can be reached than that Peter Martyr had
from Columbus’s own lips or from his manuscript jour-
nal of the second voyage (see reference to Ferdinand’s
‘‘ Historie” on p. 447), or from both, the account of the
use of the fisherman fish at the Queen’s Gardens on May
19, 1494. Consequently the first man to see and describe
the use of the sucking-fish as a living fish-hook was no
other than Christopher Columbus, the great admiral of
the ocean.

Long before this the reader has probably asked, ‘‘ What
belief is to be given these accounts of a matter apparently

452 THE AMERICAN NATURALIST [Von LIH

so incredible?’’ In answer first let us consider the innate
probability of these accounts coming from such diverse
sources. It hardly seems probable that such an extra-
ordinary phenomenon, reported separately by Dampier,
by Commerson, by Salt, by Holmwood and by Wills for
one general locality, and by Columbus and his chroniclers
for a part of the world nearly 5,000 miles away, could be
other than an actuality. Indeed Humboldt, knowing only
of Commerson’s and the Spanish accounts, gave them
full credence (1826 and 1833). He quotes Captains
Rogers and Dampier, and Columbus, and then comments
on the manner in which distant and alien peoples achieve
the same ends by diverse means, the Americans having
a fisherman-fish and the Chinese a fisherman-bird (the
cormorant), both serving the same purpose. He thinks
that the particular fish is not the small Remora but the
large Echeneis naucrates.

P. H. Gosse (1851) in the volume on ‘‘Fishes’’ in his
‘‘Natural History,’’ refers briefly to the old use of Re-
mora as a fisherman at ‘‘Hispaniola and Jamaica’’ and
concludes as follows:

From some observations of our own on the habits of a large West
Indian?! species, we are inclined to believe this account, though we do
not know that the device is at present employed.

The distinguished Cuban ichthyologist, Felipe Poey
(1856), refers to the Reversus story in a general way,
does not seem to think it improbable, but is silent as to
any such use in Cuban waters in his time, hence we may
safely conclude that the Jardinellas de la Reina no longer
witness the exploits of the fisherman fish.

But the reader may object that these stories, especially
the Columbus accounts, date back into the far past, and
may wish to know if there are any present-day statements
to be adduced confirmatory of those already given. It
may be answered that there is quite a number equally as

16 Acting on this hint, Gosse’s ‘‘A Naturalist’s Sojourn in Jamaica’’
(1851) was carefully worked over, but with negative results,

No. 628] STUDIES IN ECHENEIS OR REMORA 453

circumstantial as those quoted above. These will be taken
up chronologically for the localities involved.

And just at this point I am happy at being able to give
what is almost an eye witness account of an almost present
day use of the Remora as a living fish-hook in the very
waters in which Columbus sailed. Lady Annie Brassey
tells us that, while the ‘‘Sunbeam”’’ lay at anchor in the
roadstead of La Guayra, Venezuela in 1885:

. in one of the Indian canoes which we passed we noticed a sort of
sucking-fish (Echeneis remora), which is used in catching other fish. |
Arrived at the field of operations, the fisherman lets go an anchor and
puts the sucking-fish, attached to a long line with a buoy at the end of
it, overboard. It sees other fish at a great distance, darts after them,
and attaches itself to them by means of the sucker on top of its head.

he Indian easily raises his little anchor, paddles leisurely after the
remora, removes the captured fish into his canoe, and repeats the opera- .
tion until he has caught as many fish as he wants. Thus, one of the
ugliest and most ineapable-looking of creatures is made by savage in-
stinet to become of some use in procuring food for the superior animal.

C. F. Holder, who knew the fishes of the Florida Reef
as no other scientific man ever has, refers to Holmwood’s
accounts, makes mention of Columbus, notes that the fish
is easily tamed and goes on to say (1905) :

It is this Remora of which the story is told that fishermen employ it
in the Caribbean Sea to catch turtles. The Remora is kept, so runs the
story, in a pail; a ring is placed about its tail and to this a line. When
the men sight a turtle the Remora is slipped overboard and it is sup-
posed darts at the turtle, seizes it, and holds on with such firmness and
vigor that the animal ean be hauled in.

It is interesting to note that in the first paragraph,
Holder uses the present tense. Since he refers to them, he
certainly had knowledge of the Columbus Guiacan stories
in all of which the fish has no ring affixed to its tail, and is
carried to the fishing grounds not in a pail but adhering
to the outside of the canoe. The same account in almost
the same words is found in one of the stories in his charm-
ing little book ‘‘Stories of Animal Life’’ (1899). In this
the account of this curious fishing is somewhat amplified,
and is accompanied by a drawing, Fig. 8, Plate III, of

454 THE AMERICAN NATURALIST [ Vou. LII

this paper in a photographic reproduction of Holder’s
illustration and being a very spirited one is of interest
„and value. It is of course not a picture of an actual oc-
currence.‘

In the paper previously referred to (1905) Holder tells
of trying to catch turtles and sharks by means of a living
fish-hook, in which effort, however, he was unsuccessful.
He says:

I experimented with the Remora but the fish invariably refused to
dart after the turtle, preferring to find shelter under the boat. One
tossed to a shark was seized by the latter, that doubtless thought it a
votive offering. Possibly something was wrong: our remoras may have
been stale: they surely were not ship or turtle slayers.

In this connection the only other modern figures of
fishing with the living fish-hook may be given. Fig. 9,
Plate III, is a reproduction of one of the illustrations
from Hudson’s ‘‘Curious Bread Winners of the Deep’’
(1893). It was made to illustrate the story copied from
Ogilby’s ‘‘ America,’’ and is reproduced here for the sake
of completeness. The other figure number 10, Plate III,
is from Frederic Ober’s ‘‘Crusoe’s Island’’ (1901). He
gives the Columbus story and has’ had this figure drawn to
illustrate it. The same data without the figure is found in
an earlier book by sie et Tales of the West -
Indies,’’ 1888.

Tue Livine Fiso-Hook tx CHINESE WATERS

From the Caribbean we will go half way round the
world to find the same story in all its essentials told of the
fishermen along the southern coast of the Celestial Em-

pire. Our reference here is to Frank T. Bullen, who in —
his delightful book ‘‘Denizens of the Deep’’ (1904) gives
the following interesting account:

Turtles are many on the Chinese Coast, and the guileful Chinese
fisherman has developed a splendid plan for securing them with little

17 The same figure and essentially the same data are to be found in
Holder’s ‘‘Half Hours with Fishes, Reptiles, and Birds.’? New York,
1906, page 80 and figure 49.

' PLATE III

Fic, 8. Fishing with the living fish-hook. After Ober, 1901,
Fie. 9. Fishing with Echeneis, After Hudson, 1893.
Fic.10, Fishing with the Remora. After Holder, 1899
IG. a An Echeneis, twenty-six and one half inches lone, having a disk five
and one half inches long, lifting a bucket of water weighing twenty-four pounds.
After peice 1915

456 THE AMERICAN NATURALIST [ Von. LIII

trouble to himself. He captures some Remore, those little sharks [?]

that are so lazy that they have developed a sucking arrangement on

the top of their heads, whereby they may, and do, attach themselves to
anything that is likely to float them into the vicinity of food to be ob-

tained without effort. Carefully he welds [?] a ring round their tails in

such wise that it cannot be pulled off, and to it he attaches a thin, strong
line; then, putting out to sea with six or seven of his unwilling helpers

attached to the bottom of his sampan, he gets a good offing and waits.
patiently for the appearance of a turtle asleep upon the sea. As soon

as his keen eyes have detected one, he paddles noiselessly in that direc-
tion until, getting near enough, he ships his paddle and, with a long
bamboo, pushes off one or two or more of his Remore. Now all he
needs to do is to keep them from fastening on to the canoe again, for
they speedily discover the turtle and attach themselves to him. When

they have done so, the quaint yellow fisherman in the boat needs but to

haul in, for you may, by pulling upon a Remora from aft, tear him in

two pieces, but you cannot make him let go his hold. And so despite
his struggles the poor turtle must come [in].

In corroboration of this account, Dr. Alfred G. Mayor
tells me that he has read in Singapore newspapers that the
fishermen of that city commonly make use of the sucking-
fish in the manner just described.

FISHING WITH THE REMORA IN TORRES STRAITS

There are now to be given a number of very circum-
stantial accounts of this mode of fishing in yet another
part of the world—Torres Straits between Australia and
New Guinea. The first is from the pen of John MacGil-
livray. In his ‘‘Narrative of the Voyage of the Rattle-
snake’’ (1852), volume I, page 300, he tells of the rescue
of a white woman, Barbara Thompson by name, who had
been held captive for some years by the natives of Mura-
lug or Western Prince of Wales Island in Torres Straits
and had been named by them Giom or Gi(a)om.

In Volume II, pages 21-22, MacGillivray says:

This last (an unnamed species of turtle), I was informed by Gi’om,
is fished for in the peeing extraordinary manner. A live sucking-fish
(Echeneis remora), having previously been secured by a line passed
round the tail, is thrown into the water in certain places known to be
suitable for the purpose; the fish while swimming about makes fast by

No. 628] STUDIES IN ECHENEIS OR REMORA 457

its sucker to any turtle of this small kind which it may chance to en-
counter, and both are hauled in together.

Our next account is an eye witness one dated but a few
years after MacGillivray’s. John Jardine was for some
years police magistrate at Somerset, Cape York, where his
duties brought him into close contact with the natives. —
As a result of his experiences, in 1866 he published the
following account of fishing with the sucking-fish at Cape `
York:

A singular mode of taking the hawkbill turtle is followed by the
natives here. This custom, though said to be known so long back as the
time of the discovery of America by Columbus, is so strangely interest-
ing that I will give a short account of it as I have seen it practised. A
species of sucking-fish (Remora) is used. On the occasion to which I
allude, two of these were caught by the blacks in the small pools in a
coral reef, care being taken not to injure them. They were laid in the
bottom of a canoe, and covered over with sea-weed—a strong fishing-line

attending to the fishing lines; while I sat on a sort of stage fixed mid-
ship, supported by the outrigger-poles. The day was very calm and
warm, and the canoe was allowed to drift with the current, which runs
very strong on these shores. A small turtle was seen, and the sucking-
fish was put into the water. At first it swam lazily about, apparently
recovering the strength which it had lost by removal from its native
element; but presently it swam slowly in the direction of the turtle,
till out of sight; in a very short time the line was rapidly carried out,
there was a jerk, and the turtle was fast. The line was handled gently
for two or three minutes, the steersman causing the canoe to follow the
course of the turtle with great dexerity. It was soon exhausted and
hauled up to the canoe. It was a small turtle, weighing a little under
40 Ibs., but the sucking-fish adhered so tenaciously to it, as to raise it
from the ground, when held up by the tail, and this some time after
being taken out of the water. A strong breeze coming on, the canoe
had to seek the shore without any more sport. I have seen turtles
weighing more than 100 lbs., which have been taken in the manner
described

We next hear of this fish in Gill’s ‘‘ Life in the Southern
Isles” (1876), wherein he corroborates MacGillivray and
_ Jardine in the following citation:

458 THE AMERICAN NATURALIST [Vou. LILI

Another mode of turtling is to call in the aid of the Echeneis remora,
or sucking-fish, which is about three feet in length, and is easily caught
by a line. When caught the Straits Islanders pierce the tail, in order to
insert a strong cord, which is also wound round it for the sake of
security. Generally captive sucking-fish are kept swimming after the
canoe until a turtle is seen, when three or four of them are thrown as
near the sleeper as possible. These sucking-fishes at once attach them-
selves to the turtle, which awakes to find itself a prisoner. The cords
are now cautiously hauled in, bringing the sucking-fishes and the turtle.
This ingenious device is used only with the smaller turtle. Sucking-
fishes are sometimes kept two or three days in a lagoon or in a boat
half-filled with sea-water, until turtles are seen.

In 1888, Professor A. C. Haddon was a member of an
expedition to Torres Straits to study corals, and while
there (some eight months) he made notes of the use of
Echeneis as a turtle-catcher and of its supernatural
powers. Brief accounts of this remarkable use of the fish
were published in 1889, 1890 and 1890a (see in bibliog- |
raphy under Haddon), but as a much fuller account by
him will be given later the above need not be quoted here.

Stirred up by Haddon’s note of 1889, Sclater later in
the same year in Nature called attention to Holmwood’s
article. And, stirred by Sclater, H. Ling Roth in the
same volume of the same journal cited the account by
Ferdinand Columbus given in Churchill’s Voyages as
quoted on page 448.

Saville Kent in his book, ‘‘The Great Barrier Reef of
Australia”? (1893), has the following to say anent our
subject:

A method frequently employed by the natives of Torres Straits to
capture turtles is remarkable. The large sucking-fish, Echeneis nau-
crates, which grows to a length of three or four feet, and is distin-
guished by the natives by the title of “Gapu,” is pressed into service.
The fish is kept alive in water in the bottom of the native canoe, a thin
line being fastened round its tail and through its gills. On a turtle
being sighted in the vicinity of the canoe, the sucking-fish is thrown
towards it, and immediately swims to and fastens on its carapace. If
the turtle is of small or medium size, it is hauled in by the line, the fish
retaining its tenacious hold; but if it be a large one, a native jumps

overboard with a stronger line, and, following the smaller one down,
secures the reptile.

No. 628] STUDIES IN ECHENEIS OR REMORA 459

Corroboratory of the foregoing is the following account
extracted from Semon’s book ‘‘In the Australian Bush”’
(1899). In deseribing the catching of the turtle, Chelone
midas, by divers who jump on its back, or by fishermen
who harpoon it, Semon adds:

. . . but a third very peculiar method of capture is adopted in Torres
Straits. In clear weather and a tranquil sea, the sharp eye of the native
is able to discern any turtle reposing on the bottom of the sea in the
neighborhood of the coral reefs. Now a sucking-fish, or Echeneis, to the
hind fin of which a long string has been fixed, is thrown into the water
above the place where the turtle has been seen. It will immediately
descend into the depth and attach itself to the shell of the reposing
Chelonian, and as a communication is thus established between the boat
and the turtle, a native following the leading string, dives and winds a
rope round the beast, as the sucking-fish does not attach itself quite
firmly enough for the fisherman to draw the heavy weight up by it.15

This last statement must not be interpreted as contra-
dictory of the foregoing accounts of catching turtles on
the surface with the sucking fish. Bringing boat and
turtle together on the surface by pulling on the line is one:
thing, hauling a turtle up from the bottom is quite an-
other; as any reader knows who has ever endeavored to
land a ray or other large flat fish which insisted on cling-
ing to the bottom. This latter is purely a problem in
hydrostatics.

Entirely independent of any of the foregoing accounts
is that of the Australian ethnologist, W. E. Roth. Here
the location (Tulley River) is different, as is the final
manner of taking the fish, turtle, or dugong. Roth’s
statement follows:

On the coast-line in the neighborhood of the Tulley River, the sucker-
fish, Remora, is utilised as a guide for spearing or harpooning fish, as
well as turtle and dugong. This sucker-fish, known to the Mallanpara
blacks as kamai, is found usually on the rocks at the outlying arak
and sometimes stuck on their own canoes. It is removed, kept in
canoe, bark-trough, o with a little water, and left there for a Tew
days. Then, going out to sea, the native ties a fine twine round the
Remora’s tail, and as soon as he sights any big fish, turtle or dugong,

18 This account is also found in the German edition of Semon’s book
published at Leipzig in 1903.

460 THE AMERICAN NATURALIST [ Von. LIII

advances his canoe as far as possible, and drops the sucker-fish over-
board. In all probability, the sucker will go straight for the object and
attach itself: it acts only as a guide, and tells the hunter the next move
of his prey. The aboriginal now plays the line out very guardedly,
draws it in with equal care and caution, and as soon as the length sub-
merged reaches a point on the line, previously marked, he knows that
he is within striking distance, and as his quarry comes to the surface,
uses the spear or harpoon accordingly. It must be borne in mind that
in no sense does the sucker-fish pull the prey into the hands of the
hunter: it only indicates the direction in which the harpoon, ete., can be
advantageously thrown.

The account given by N. W. Thomas in his book
‘‘ Natives of Australia’’ (1906) is taken almost verbatim
from the above and beyond this mere citation no notice
will be taken of it here.

We now come to another account of the peculiar use of
Echeneis under discussion, and I am able to offer no less
an authority than the ‘‘ Encyclopædia Britannica,’’ in the
eleventh edition of which, in Volume XXII (1911), in the
article on Queensland, Australia, Mr. T. A. Coghlan
writes:

‘In Torres Strait and the northern coast the hawksbill turtle .. . is
said to be captured in a peculiar manner, the sucking-fish or remora
(Echeneis naucrates) being utilized by the islanders for that purpose.
The remora is carried alive in the bottom of the canoe, a long thin line
being attached to the fish’s tail and another usually to the gills. Ona
turtle being sighted and approached to within the length of the line, the
sucking-fish is thrown towards it, and immediately swims to and attaches
itself by its singular head sucker to the under surface of the turtle
which if of moderate size is easily pulled into the canoe.

During the year 1898, Professor A. C. Haddon was
leader of the Cambridge University Anthropological Expe-
dition to Torres Straits. On this expedition he made an
extensive study of the use of the fisherman fish. Professor
Haddon’s data is so complete that he has effectually
settled the matter of the present-day actual use of the fish
for taking other fish, and since his reports are of the high-
est value, putting as they do the imprimatur of truth on
the whole matter, they will be referred to in some detail.

Professor Haddon’s first account based on the data of

i

No. 628] STUDIES IN ECHENEIS OR REMORA 461

his second expedition is to be found in his ‘‘ Head Hunters:
Black, White and Brown’’ (1901). This gives essentially
the same data as that contained in the short article in
Folklore, 1890, but for fuller accounts we must turn to the
various reports of the Cambridge a Expe-
dition to Torres Straits.

Taking up these reports chronologically fits in well with
the scheme of this paper, as will be seen presently.
Volume V (1904) deais with the ‘‘Sociology, Magic and
Religion of the Western Islanders.” Here Haddon gives
three folk tales, one of which has to do with the origin of
the use of the Gapu (the native name of the sucking-fish),
and two with itsuse. Later in the same volume Dr. Rivers
gives a very detailed account of the method of procedure
in fishing with the Gapu. This data will be found later in
Haddon’s final account of the use of this fish. Further
along in Volume V Haddon and Rivers give accounts of
the Gapu as a totem.

Volume VI of the Reports bearing date 1908 has for its
title the ‘‘Sociology, Magic and Relation of the Eastern
Islanders.’? These peoples do not seem to have so many
tales of the Gapu as their western brethren since Haddon
records but two. It seems apart from the purpose of this
paper to insert any of these folk tales here, but it is my
purpose later with Professor Haddon’s kind permission
to collect them and publish them in a short article.

We now come to the latest, most detailed, and most valu-
able of all the accounts of the use of the living fish hook in
Torres Straits. In Volume IV of the Reports issued in
1912, Professor Haddon gives a very circumstantial ac-
count and this will be quoted in full. In this volume,
dealing with arts and crafts, fishing with the sucking-fish
is frequently referred to. The fish is well known to the
natives as their myths and legends show and it is a com-
mon motif in their ornaments and ornamentation. Had-
don’s account of its use now follows:

The most interesting method of catching turtle is that in which the
sucking fish (called gapu in the western part of the straits, and gep in

462 THE AMERICAN NATURALIST [Vou. LIII

the eastern) is employed. . . . The sucker-fish is not used to haul in the
large green turtles; I was repeatedly assured that it would be pulled off,
as the turtle was too heavy; but small ones are caught in this manner.

According to one of thé folk tales, there was a time when the pani
of Badu did not know how to catch turtle by means of the sucker-fish,
and they used to employ a black toothless “ dog-fish,” Kumsar, when
they went for turtle. The story goes on to tell how Bia taught his
fellow islanders how to employ the sucker-fish. In the Bomai-malu
legend of the Miriam, it is stated that Barat of Moa, according to the
fashion of olden times, tied a rope around the tail of a kamosar, then
he made a sucker-fish, and instructed the Western Islanders who were
with him how to catch turtle with it. I do not understand how turtle
could be caught by a “ dog-fish,” but as the identity of this fish, which
is said to live in the crevices of the rock in deep water, is unknown,
nothing further ean be said, except to hazard the suggestion that it may
be an unidentified kind of lamprey; but against this it must be stated
that no member of the Cyclostomata is known from Queensland waters,
though Mordacia mordax occurs in Tasmania and species of Geotria are
found in southern Australian waters.

I was informed that in leashing a sucker-fish, a hole is made at the
base of the tail-fin by means of a turtle-bone and one end of a very
long piece of string inserted through the hole and made fast to the tail,
the other end being permanently retained. A short piece of string is
passed through the mouth and out at the gills, thus securing the head
end. By means of these two strings the fish is retained, while slung
over the sides of the canoe, in the water. The short piece is pulled out
‘of the mouth of the fish when the turtle is sighted and the gapu is free
to attach itself to the turtle.

According to Professor Haddon there is a certain cere-
monial or set rule of procedure always definitely followed
in fishing with the gapu. This he describes as follows, his
data being chiefly taken from Rivers as noted above:

When starting on a trip to fish for turtle by means of the sucker-fish,
the owner (or captain) of the canoe gives the order where to go and
when to let go the anchor, having arrived at their destination.

The buai-garka (mate, also brother-in-law of the owner or captain)

makes a fire on which he places some turtle-bone which the owner has
brought with him. When the bone is charred the buai-garka breaks it
up and throws it into the water so as to attract the sucker-fish. When
one is caught it is the duty of the buai-garka to attach to the fish the
leashing which he had previously made.

The direction of affairs is now assumed by the buai-garka, who gives
the word to move to another place, and the directions where to go.

No. 628] STUDIES IN ECHENEIS OR REMORA 463

When he gives the order to stop, the mat sail is rolled up by the other
men (or at the present time the sail is lowered), he not taking any part.
He gives the order to paddle till he sees the turtle, then gives the word
to stop, and the anchor is let go by the owner, having been previously
shifted to the stern of the canoe. When the buai-garka sights a turtle
swimming deep down in the water, he removes the mouth string from
the sucker-fish and throws the fish overboard with the tail-line attached
and plenty of slack is thrown with it, he then hauls in the superfluous
slack and as far as possible indicates the direction of the turtle by pres-
sure on the line. The sucker-fish on perceiving the turtle immediately
swims towards it, and attaches itself to the reptile’s carapace. When
this is accomplished, the buai-garka gives the order to heave up the
anchor and move the boat up to the position of the turtle.

One of the crew (but not the buai-garka), with a long rope attached
to the right upper arm, dives into the water, and is guided to the turtle
by the line fastened to the fish’s tail. On reaching the turtle, the man
gets on to its back and passes his arms behind and below the fore
flappers and his legs in front of and below the hind flappers, or
secures it in some other way. The man is then rapidly drawn up to
the surface of the water bearing the turtle with him. On the arrival of
the diver the sucker-fish usually shifts its position from the upper to
the under surface of the turtle. As soon-as enough turtle have been
obtained, the owner of the canoe gives the order to go home, and the
buai-garka resumes his subordinate functions, and resigns into the hands
of his brother-in-law the direction of affairs which had been his part
during the actual process of fishing. The bwai-garka knows whether
the fish has attached itself to a turtle or to a shark by the nature of the
motion of the string. If the pull is intermittent it means that the fish
has adhered to a shark, but if steady, then a turtle has been secured.

Ina footnote Professor Haddon tells us that the sucker-
fish is eaten at the end of the day’s fishing. This seems
like a very wasteful course of action, but it may have
arisen because of the difficulty in keeping the fish alive
until the next time it would be needed. In text-figure 5,
we have a native drawing showing how the ‘‘Gapu’’ is
attached to the canoe during the trip to the fishing grounds.

We have in Holmwood’s account a description with
figures of how the leashing is accomplished, and Haddon
also is too good a scientist to leave us in doubt as to how
the Torres Straits natives manage this matter. In addi-
tion to what has already been given as to the manner of
making the leashings fast, his detailed account is as
follows:

464 THE AMERICAN NATURALIST [ Vou. LIII

In order to understand the method of leashing a sucker-fish, I induced
a native to make a model of a gapu for me. Fig. 173 [present text-
figure 5] indicates diagrammatically the arrangements. A loop is in-
serted by means of a wooden arrow point through the gills and out at the
mouth, the ends are passed through the loop, and one of the strands is
threaded through the other, the two are then twisted into a string.
The mouth string is called gudaz and is made of the inner bark of the
root of the wali tree. The other end of the gudaz is tied into a slip
knot, kaza wiaikab, the end of a long piece of twine is simply bent twice
round the string at the knot; when the free end of the gudaz is pulled

TEXT-FIGURE 5. Showing how the sucker-fish is leashed by the natives of Torres
nebo Drawing of Tagai and Kareg in their canoe by Gizu of Mabuiag, reduced
by one half. In this drawing, the canoe, Kareg and the gie fish are repre-
pon the wrong way round. After Haddon, 1912.

the knot runs out, and the twine (one end of which is still held, by the
fisherman) slips off the gudaz. The main fishing string is a very long
and strong cord of twisted coco-nut fibre, igal; this is fastened to a braid
of plaited string, dan, the other end of which is bent round on itself so
as to form a loop; the end of the dan and the loop are bound round with
wali. The loop is furnished with two strings of wali; it looks as if
these were threaded through the tail of the fish above and below the
vertebral column and tied together on the other side. Another lashing
binds the cord close to the side of the narrow portion of the tail.

-a

No. 628] STUDIES IN ECHENEIS OR REMORA 465

The sucker-fish is so well known to the natives as to
give rise to a decorative motive in their decorative art.
Haddon gives numerous figures of this. It is also a sub-
sidiary totem of one of the clans in the western islands
and Haddon thinks may once have been a chief totem of a

TEXT-FIGURE 6. Native drawing illustrating the method of attachment of the
sucker-fish to the canoe in Torres Straits.

larger but now extinct clan. The natives moreover ascribe
to it considerable intelligence. Haddon thus concludes
his interesting account:

The natives have great respect for the sucker-fish and firmly believe
it to possess ominous For example: when the fish does not

part of the canoe is not secure; when there is something the matter
with the bow of the canoe, the fish is said to attach itself to the neck of
the turtle, but should the stern of the canoe be weak, the fish adheres to
the extreme hinder end of the carapace; when it fixes itself firmly to
the front part of the carapace, the canoe is strong; when it goes to one
side of the carapace or keeps moving about, it shows that the leashings
of the float to the outrigger on that particular side are insecure. More
than once I was told, Gapu savey all same man. I think him half devil
(i. e., spirit) .29

One other account is now to follow and all the known
data will have been fully presented to the reader. E. J.
Banfield, the well-known Australian, lived for a number
of years on Dunk Island, off the coast of tropical Queens-
land, in about latitude 18° S. His experiences there are

19 After Haddon’s full and very detailed "on Meek ’s brief reference
(1913) that the ingenious natives of Thursday Island ar the adjacent
parts of Torres Straits use one fish to catch another—i. e., the sucking-fish
with a string fastened to itstail—needs no fuller Aaoi than that given
in this footnote.

466 THE AMERICAN NATURALIST [ Vou. LIII

recounted in a charming book, ‘‘The Confessions of a
Beachcomber.” This was first published in 1908, and re-
printed in 1910, and again in 1913. The following quota-
tions are from the 1913 reprint:

Generally unprogressive and uninyentive, the aboriginals of the coast
of North Queensland apply practically the result of the observation of a
certain fact in the life history of a fish to obtain food. By them the
sucker (remora) is not regarded as an interesting example of a fish... ,
but as a ready means of effecting the capture of ... two... animals
(turtle and dugong), always eagerly hunted for their flesh.

Other countries have sucker-fish of different form; but it remained
for the benighted Australian blacks, among a few other savage races, to
make practical use of the creature, which, as a means of locomotion,
forms strong attachments to the dugong, turtle, shark and porpoise.
It can hardly be called domesticated, yet it is employed after the manner
_ of the falcon in hawking, save that the sucker is fastened to a light line
when the game is revealed.

Having located the haunts of a remora the blacks feed
it from day to day until its shyness is worn off, and then
catch it with a hook.

Having secured the sucker, the blacks farm it in their haphazard
fashion. They fasten a line above the forked tail so securely that it
eannot slip, nor be likely to readily eut through the skin, and tether it in
shallow water, when it usually attaches itself to the bottom of the canoe.
When as the result of frequent use and heavy strain, the tail of the
sucker is so deeply cut by the line that it is in danger of being com-
pletely severed, a hole is callously bored right through the body beside
the backbone, and the line passed through it for additional security.

When ready to hunt for turtle, the natives armed with
spears go out in their bark canoe to the bottom of which
one of the sucking fish is attached by its sucker. When
they reach a locality where turtle abound, they soon get
into action.

In sight of the game the sucker which has been adhering to the bottom
of the eanoe is tugged off and thrown in its direction. As a preliminary
the dise and shoulders of the sucker are rigorously serubbed with dry
sand or the palm of the hand, to remove the slime and to excite the
ruling passion of the fish. It makes a dash for a more congenial com-
panionship than an insipid canoe. The line by which it is secured is
made from the bark of the “Boo-bah” (Ficus fasciculata) and is of

No. 628] STUDIES IN ECHENEIS OR REMORA 467

two strands, so light as not to seriously encumber the sucker, and yet
strong enough to withstand a considerable strain. Two small loops are
made in the line at intervals of two fathoms from the sucker, to act as
indicators.

As soon as the sucker has attached itself to the turtle, a slight pull
is given and the startled turtle makes a rush, the line being eased out
smartly. Then sport of the kind that a salmon-fisher enjoys when he
has hooked a 40-pounder begins. The turtle goes as he please; but when
he begins to tire, he finds that there is a certain check upon him—slow,
steady, never-ceasing. After ten minutes or so a critical phase of the
sport occurs. The turtle bobs up to the surface for a gulp of air, and
should he catch sight of the occupants of the canoe, his start and sudden
descent may result in such a severe tug that the sucker may be divoreed.
But the blacks watch, and in their experience judge to a nicety when and
where the turtle may rise; telegrams along the line from the sucker give
precise information. They crouch low on their knees in the canoe, as
the game emerges with half-shut eyes and dives again without having

rocks and coral, and endeavors to free himself from the sucker by rub-
bing against the boulders. Knowing all the wiles and maneuvres, the
blacks play the game accordingly, and hour after hour may pass, they
giving and taking line with fine skill and the utmost patience. The
turtle has become accustomed to the ineumbrance and visits the surface
oftener for air. One of the harpoons is raised, and as the turtle gleams
grey, a couple of fathoms or so under the water, the canoe is smartly
paddled towards the spot whence it will emerge, and before it can get
a mouthful of air the barbed point, with a strong line attached, is
sticking a couple of inches deep in its shoulder.

From the foregoing interesting accounts it is clear first
that in Torres Straits at the present time the aborigines
use the sucking fish as a living fish-hook just as fishermen
in other regions use a veritable fish-hook, i. e., to bring
the fish or turtle to the gaff. The gaff is, in the last ac-
count quoted, a harpoon or spear, according to others a
native who dives down, guided to the turtle by the line
attached to the Echeneis, and who then ties a line to a
flipper of the turtle.

THE ENGLISH SPARROW HAS ARRIVED IN DEATH
VALLEY: AN EXPERIMENT IN NATURE?

THe English sparrow first became well established in the
United States in 1860-1864 in the vicinity of New York City.
Several small plants had been made in other Atlantic cities
within the few years preceding, but practically all of these are
definitely known to have failed. The original stock is in nearly -
all the cases of importation known to have been obtained in Eng-
land. Its spread through the eastern United States after once
established was phenomenal; its rate of invasion towards the
west only slowed up at about the 100th meridian, and this, sig-
nificantly enough, is about at the line limiting a great many spe-
cies of native eastern birds toward the west and of native western
birds toward the east. Nevertheless, the English sparrow ulti-
mately crossed this barrier, constituted by change in humidity,
and it has continued expanding its range until it exists now in
nearly every part of every state in the Union. It has also ex-
tended throughout southern Canada and has become well settled
in the Hawaiian Islands.

In California the English sparrow was first noticed in 1871 or
1872, in San Francisco, and it quickly thereafter appeared in
many of the towns in the west-central part of the state. But it
was very slow to enter southern California. It did not reach
Los Angeles for nearly thirty-five years, in 1907; and San Diego
was not reached until 1913. To-day it is familiar in practically
every town ‘‘south of Tehachapi.’’ Among the places in Cali-
fornia now inhabited by the English sparrow, to designate some
of those showing extremes of climate as regards temperature
and humidity, are Brawley, Imperial County, and Sisson, Sis-
kiyou County; Needles, San Bernardino County, and Eureka,
Humboldt County.

In 1917 the California Museum of Vertebrate Zoology under-
took as field work for that year a study of the vertebrate animal
life of the Inyo region of southeastern California. In connec-
tion with this work it was the writer’s not unpleasant fortune to
spend the month of April in Death Valley. What was his sur-

1 Contribution from the Museum of Vertebrate Zoology of the University
of California.

468

No. 627] SHORTER ARTICLES AND DISCUSSION 469

prise to find there a thriving colony of English sparrows. These
were established on the Greenland Ranch (otherwise known as
Furnace Creek Ranch), elevation 178 feet below sea level.
Specimens were collected, both as alcoholics and as dry study
skins, but not to an extent to threaten the persistence of the
colony. For here, it occurred to the writer, we had at hand a
particularly convincing ‘‘experiment’’ already under way, of
just the sort called for by certain critics of the work of the sys-
tematist and distributionist, which in time would test the ques-
tion of the evanescence versus the relative permanence of char-
acters of the category commonly viewed as subspecific.

The sparrows of Furnace Creek Ranch, which were estimated
to number about fifty, had their main headquarters in the tops
of the several tall Washington Palms which overshadow the
ranch house; also several nests were seen in the Fremont cotton-
woods which line the irrigation ditches along the alfalfa fields
for a quarter of a mile down toward the glistening borax flats.
The traveller on entering Death Valley is impressed by Green-
land Ranch as a wonderfully rich oasis surrounded by a desert
of surpassing barrenness. The English sparrow colony there is,
then, isolated under a climate that is probably of the greatest ex-
treme in the direction of high temperature combined with low
relative humidity, of any place in North America.

Greenland Ranch is owned by the Pacific Coast Borax Com-
pany, who value it for its output of alfalfa hay and for certain
appurtenant water rights, there being a constant flow of forty
inches from the warm springs nearby. Fortunately for our
present problem, the company has for years required its mana-
gers to keep a daily record of weather conditions. There is a
standard instrument shelter, and the records are kept in avail-
able form, and furthermore have been transmitted regularly to
the United States Weather Bureau. Without going into details
here, it is of interest to note that the highest recorded tempera-
ture for any place in the United States was observed there on
July 10, 1913, when an afternoon temperature of 134° Fahren-
heit in the shade was reached.

As to the time of appearance of the sparrows in Death Valley
I have good reason to rely on the statements of Mr. Oscar Den-
ton, who is the present manager of the Greenland Ranch. He
Says that he first saw them in the ranch yard five years ago
(1914). That was about the time the Death Valley spur of the
Tonopah and Tidewater Railroad was run to the present location

470 THE AMERICAN NATURALIST [Vou. LIL

of Ryan. Ryan, by the way, is the terminus of the narrow-
gauge line, wherever that terminus happens to be, and this shifts
about as determined by the extent of the different ledges of
borax ore mined. The borax deposits on the floor of Death Val-
ley are no longer gathered. The day of the 20-mule-team borax
wagons is gone except on the labels. It is cheaper to handle
the richer borax ore high on the mountain sides and to reach
these ledges by railroad. The present Ryan, the nearest the
railroad has so far gotten to Death Valley, is 17 miles from
Greenland Ranch and 3,000 feet altitudinally above it. I saw
English sparrows there repeatedly in April and May, 1917, as
also at Death Valley Junction, 40 miles farther away, on the
Tonopah and Tidewater Railway. Mr. Denton believes, and I
think he is likely right, that the sparrows followed the construc-
tion camps along the route of the T. & T. R. R. from Ludlow to
Death Valley Junction and thence along the narrow-gauge to
Ryan. It may be further suggested that since hauling is done
from time to time down the 17 miles of Furnace Creek Wash
from Ryan to Greenland Ranch this is the route probably trav-
elled by those sparrows which reached Death Valley. It is less
probable to my mind that the birds simply started out overland,
from some more distant point, and a pair or more just happened
to reach this remote and forbidding valley. It is true, however,
that the green of the ranch shines out conspicuously for miles
round about and would surely attract to it any vagrant sparrow
coming within sight.

We here in America have been accustomed to think of the
- English sparrow as a full species, Passer domesticus. The bird
was originally named by Linnaeus, and thus has seemed from all
standpoints to constitute a truly ‘‘ Linnaean species.’’ How-
ever, recent developments in the geographic knowledge of birds
in the Old World has brought out the fact of geographic varia-
tion within the species Passer domesticus as previously under-
stood, and also that a number of forms once considered specifi-
eally distinct are really connected with the domesticus stock
through ordinary geographic intergradation. Hartert (1910,
pp. 147-151) after a study of the group came to recognize no
less than eight subspecies occupying different areas in Europe,
western Asia and northern Africa. Subsequently, at least two
more races have been named. And now, a German, Klein-
schmidt, has discovered that the sparrows of England are distin-
guishable from those on the continent. The latter, having been

No. 627] SHORTER ARTICLES AND DISCUSSION 471

the basis of Linnaeus’s name, becomes Passer domesticus domes-
ticus, and the sparrow of England Kleinschmidt mames (cited
under date 1915, though I have not seen the original description
myself) Passer domesticus hostilis. As pointed out by Ober-
holser (1917, p. 329), since the American stock came from Eng-
land our bird must-also be known under this name. And fur-
thermore, the vernacular term, European house sparrow, which
some people have preferred because of a fancied unpleasant
association in the name English sparrow, can not be used prop-
erly for the American bird.

The point I wish to make now is that the English sparrow,
which is spread all over the United States, is itself a subspecies
of a wide ranging and decidedly variable species which is thus,
geographically speaking, quite like our American song sparrow,
or the horned lark. In the Old World, each race ‘‘stays put”’
as regards aggregate of population, each in its own faunal area
just as do our own song sparrows. All of these races are non-
migratory. Passer domesticus hostilis Kleinschmidt is also non-
migratory, as far as I have been able to learn, wherever it now
occurs, north and south, in America. But here, by reason of its
marvellous powers of accommodation, and finding no competitor
in exactly its own ecologic niche, it has gradually advanced its
frontiers and overleaped all the faunal boundaries which hem in
the habitats of our native bird races; and we find flourishing
representations of it under the most diverse conditions of envi-
ronment, as for example those shown in contrast by Death Val-
ley and Boston.

Possibly our critics have been merely baiting us when they
asked us to transplant a desert song sparrow to the humid coast
belt and ‘‘see what would happen.’’ But is not this demand
met exactly in the case of the English sparrow, only in reverse
direction? I have carefully compared the seven skins taken in
Death Valley with others taken in Berkeley, and also with ex-
amples taken in the eastern United States, without finding any
peculiarities of color tone, extent of markings, or dimensions.
And I think my eyes are pretty well trained to find small sub-
specific distinctions, at least of such magnitude as characterize
the currently recognized subspecies of song sparrows, Savannah
sparrows, and horned larks. The Death Valley birds, it is true,
stand out rather sharply from most of the material taken else-
where, but only in that they are fresh and clean, and lack the-

* sooty overcast of the majority of town birds. To repeat, no dif-

472 THE AMERICAN NATURALIST [Vou. LIII

ferences are now discernible from place to place in North Amer-
ica, in so far as perfectly comparable material is at hand. This
accords with the findings of Phillips (1915), which also were
practically negative.

Are we not to infer, then, that there has not as yet been suffi-
cient time (up to three years and as many possible generations
in Death Valley and up to sixty years elsewhere in North Amer-
ica) for the impress of diverse environments in the different
parts of the territory newly occupied by Passer domesticus hos-
tilis to bring physical changes in the birds of sufficient magni-
tude for the modern systematist to detect? Is there not here a
demonstration of the relative permanence of subspecific charac-
ters far beyond what many naturalists have supposed? Are not
such characters in general far more likely to be germinal than
somatic ?

How intensely interesting it will be to watch the course of this
‘‘experiment,’’ now under way, irrespective of human effort, in
Death Valley, with ‘‘controls’’ vigorously maintaining them-
selves (against man’s wish!) in San Diego, Berkeley and Boston.

But perhaps it will be urged that the conditions of an orthodox
experiment are not here properly met. The ‘‘factors’’ of the
environment are not sorted out, and none is under any kind of
regulation. Moreover, rigid control has not been secured, in
that there is no way in which any of the naturally established
colonies of English sparrows can be strictly isolated and kept
from genetic contamination by new influxes of birds from else-
where.

In reply, I would say that we are not expecting more from our
natural experiment than the demonstration of what we set out
to prove, namely, the length of time necessary for the develop-
ment, in a stock under natural conditions some of which are
known, of characters of subspecific value. In the breeding cage
there are always ‘‘unknown’’ factors; so let us admit the exist-
ence of those in the wild as not invalidating the ‘‘experiment”’
as such. In nature, subspecies have differentiated under just
the conditions self-imposed by the English sparrows through
their powers of invasion. Individual song sparrows and horned
larks are continually overstepping the bounds of the habitats of
the races to which they belong and have doubtless done so since
the initiation of their respective descent lines. But differentia-
tion of the mass has taken place, wnder just these conditions.

OSEPH GRINNELL

| THE
AMERICAN NATURALIST

Vou. LIII. » November—December, 1919 No. 629

INHERITANCE OF WHITE-SPOTTING AND
OTHER COLOR CHARACTERS IN CATS

DR. P. W. WHITING

FRANKLIN AND MARSHALL COLLEGE, LANCASTER, PA.

In a previous paper! I have presented data bearing on
the general subject of the inheritance of coat-color in cats.
The experiments at the University of Pennsylvania were
still in progress when the paper was published and
further results have since been obtained. Unfortunately
an eczema infected the stock and the investigations were
brought to an end by the death of several animals. It is
thought advisable therefore to present the rest of the
data in the present paper and to summarize results thus
far obtained.

Numbers denoting individuals or matings are inserted
as in the previous paper for the purpose of cross ref-
erence. In the genetic formule A’ denotes much-ticked ;
A, little-ticked ; a,non-ticked; B’ denotes lined; B, striped;
b, blotched; M, denotes intensely pigmented; m, maltese
dilution; W denotes dominant solid white; w, color; Y de-
notes yellow; y, black; Yy, tortoiseshell. Symbols are
omitted when character was for any reason undetermined.

A cream male (24) (b.m.w.Y) was crossed (48) to a
solid-white yellow-eyed half sister (29) (W) from mating
43 (a white male by an ‘‘anomalous’’ (Yy) cream female
(23) mentioned below). There were produced one orange
male (b.M.w.Y), and one cream female (b.m.w.Y). Both

1 Whiting, P. W., ‘‘Inheritance of Coat-Color in Cats,’’ The Journal of
Experimental Zoology, Vol. 25, No. 2, April, 1918.

474 THE AMERICAN NATURALIST (Vot. LI

had an extreme amount of white-spotting. The same
male (24), when crossed (51) to his ‘‘anomalous”’ (Yy)
cream mother, (23) (b.m.w.Yy), sired two cream females
(b.m.w.Y). He was also crossed (53) to a black female,
(32) (a.B.M), and sired one maltese male, (a.B.m.w.y)
and four tortoiseshell females, (one a.B.M, one b.m, and
two a.M). :

The progeny from the ‘‘anomalous’’ (Yy) cream female
(23) by her cream son (24) now consists of four litters
(41, 46, 50, and 51) containing two maltese males (yX
—), one cream male (YX —) and four cream females ©
(YX IX}.

There are now in all eight matings of ‘‘yellow’’ male
(YX —) by ‘‘black’’ female (yX yX) giving sixteen
‘black’? males (yX —) and seventeen ‘‘tortoiseshell’’
females (YX yX).

Dr. Charles Penrose, of Philadelphia, very kindly
loaned his Caffer cat for crossing, a much-ticked lined
male (21) (A’.B’.M.y) mentioned in the previous paper.
A cross (52) made with an orange striped female (31)
(B.M.Y), from mating 37 produced three tortoiseshell
females,—a much-ticked lined (A’.B’.M.Yy), a lined with
ticking present but with so much yellow that the exact
degree was uncertain, (B’.M.Yy), a ticked of uncertain
degree in which the banding was also uncertain on ac-
count of admixture of black and yellow, (M.Yy).

The same male (21), crossed (55) to an orange striped .
sister (33) (B.M.Y) of female 31, sired five orange lined
males (B’.M.Y). —

The same male (21) crossed (54) to a tortoiseshell (28)
(a.B.M.Yy) sired four lined non-yellow kittens,—two
little-ticked males (A.B’.M.y), and two much-ticked
females (A’.B’.M.y).

The same male (21) crossed (56) to a blotched maltese
tortoiseshell (13) (A.b.m.Yy) sired four lined orange
males (B’.M.Y) and two lined tortoiseshell females
(B’.M.Yy).

When he was crossed (57) to a dilute tortoiseshell (34)

No. 629] INHERITANCE OF WHITE SPOTTING 475

(a.b.m.yY), there were produced a lined orange male
(B’M.Y) and a lined tortoiseshell female (B’.M.Yy).

When he was crossed (58) to a tortoiseshell female
(30) (a.B.M.Yy) he sired two orange lined males
(B’.M.Y), one little-ticked, lined male (A.B’.M.y), and
two lined females,—one tortoiseshell with so much yellow
that degree of ticking could not be made out (B’.M.Yy),
and one non-yellow with so much white that degree of
ticking could not be made out (B’.M.y).

The crosses of this Caffer cat are reciprocal to those
summarized above, crosses of ‘‘yellow’’ males (YX —)
by ‘‘black’’ (yX yX) and by ‘‘tortoiseshell’’? (YX YX)
females. Here we have a ‘‘black’’ male (yX —) by ‘‘yel-
low” females (YX YX) giving five ‘‘yellow’’ males
(YX —) and three ‘‘tortoiseshell’’ females (YX yX) and
a ‘‘black’’ male (yX —) by ‘‘tortoiseshell’’ females (YX
yX) giving three ‘‘black’’ males (yX —) and seven ‘‘yel-
low” males (YX —) and four ‘‘tortoiseshell’’ females
(YX yX) and three ‘‘black’’ females (yX yX). It may
be seen, therefore, that the principle of sex-linkage ap-
plies in all these cases.

The progeny of the Caffer cat (21) are of interest also
from the point of view of ticking and banding. The es- _
sential characteristic of Caffer is the narrow banded or
‘‘lined’’ condition. Banding of intermediate width,
‘*striping,’’ acts as a recessive as previously shown and
the widest bands, ‘‘blotches,’’ are recessive to both
‘*lines’’ and ‘‘stripes.’’ The parents (18, 19) of this cat
were both lined, but produced blotched offspring as well
as lined. Evidently this cat (21) is the homozygous seg-
regate, for twenty-four of his twenty-five offspring are
: certainly lined and in the other (52.3) there is so much
white-spotting and so much intermixture of yellow and
black in the pigmented areas that the condition of band-
ing is uncertain. It is probable, however, that even in
this case a wider type of bands,—stripes or blotches,
would have been more easily seen. Of the females to
which cat 21 was crossed, two (13, 34) were blotched, and

476 THE AMERICAN NATURALIST (Vou. LIII

the other four (28, 30, 31, 33) were striped but known to
be carrying blotched. Lined is therefore dominant to
both striped and blotched as previously stated. The re-
sults thus far obtained do not demonstrate the allelomor-
phism of the three types of banding. In order to do that
it would be necessary to cross one of the offspring carry-
ing striped (B’B) to blotched cats (bb). All kittens
should be lined (B’b) or striped (Bb). If blotched oc-
curred it would demonstrate that two loci were involved,
Ll and Ss. Blotched would then be ll.ss, and the nomen-
clature would have to be changed.

The production of orange and tortoiseshell lined cats is
of interest. They are as expected in every way com-
parable to other oranges and tortoiseshells except for the
narrower bands.

Results in regard to ticking may now be considered. In
the previous paper a was used to denote lack of ticking;
A, little-ticked or dark tabby; and A’, much-ticked or
light tabby. It now appears that there are two hered-
itary grades of ticking previously grouped under A’ be-
tween which there is a fairly wide difference. Compar-
ison of kittens at birth or of adult cats makes the distinc-
tion clear. During growth intergradations appear for
ticking increases with age as in rodents. A’ should
therefore be divided into A’, extreme-ticking, and A”,
much-ticking. Fully as much difference exists between
A’ and A” as between A” and A.

A blotched male (11) crossed (14) with a black female
(15) sired four blotched kittens, and a blotched female
(14) crossed (31) with a lined little-ticked male (18) pro-
duced one lined and three blotched offspring. The two
blotched cats (11 and 14) were extremely-ticked, A’, as
were also the eight kittens. A much-ticked, A”, Caffer
female (19) when crossed (19) to the little-ticked, A,
Caffer male (18) produced one much-ticked (21) and
three little-ticked. The much-ticked mother (19) and
son (21) are very similar and contrast strongly with the
extremely-ticked cats mentioned as well as with little-

No. 629] INHERITANCE OF WHITE SPOTTING 477

ticked. With the exception of five kittens, the offspring
of the much-ticked male (21) are useless for determining
degree of ticking on account of the presence of yellow.
Three kittens (58.2, 54.3, and 54.4) are little-ticked like
their grandfather (18). Two kittens (54.1, 54.2) were
much-ticked like their father and grandmother. The
mothers of all of these kittens were non-ticked. The same
degrees of ticking, A” and A, have been possessed by
three generations.

Skins illustrating the three types of ticking A’, A”, and
A are preserved for reference.

The crosses summarized in the preceding paper and
above may now be considered from the point of view of
white-spotting. Solid-white acts as a complete dominant
to other colors as shown in the previous paper. White-
spotting as seen among cats in general grades all the way
from solid-white to self. In individual fraternities, how-
ever, it may show wide and clean segregation as the
crosses below demonstrate. A ‘‘self’’? cat may have a
minute white spot on breast of belly or a few sparsely
scattered white hairs. In this case it might be called
near-self. Restricted spotting denotes white on nose,
breast, belly, or feet. It segregates widely from near-self
in the crosses here considered, but grades into moderate
spotting, which denotes the further extension of white to
sides of body as well. Moderate spotting in turn grades
into considerable, which denotes more white than color.
Extreme spotting denotes that pigment is limited to
small spots on head, back, or tail.

Crosses involving only self, restricted and moderate
spotting, and solid-white may be considered first.

A self male (18) crossed (19, 31) to two self females
(19, 14) sired eight self. One of these self offspring (21)
crossed (55) to a self female (33) sired five self. Self
may therefore breed true.

The first mentioned self male (18) crossed (28, 29)
with two restricted spot females (10, 2) sired three self
and four restricted spot. The other self male (21)

ka

478 THE AMERICAN NATURALIST [Von. LIIT

crossed (54, 56) to two restricted spot females (28, 13)
sired five self and five restricted spot. A restricted spot
male (8) crossed twice (9, 30) to a self female (20) sired
three self and seven restricted spot. A restricted spot
female (3) crossed (16) to a self male (6) produced six
self. Self by restricted spot therefore has produced sev-
enteen self and sixteen restricted spot, the expectation
if restricted spot is heterozygous.

The restricted spot male (8) crossed (12, 33) to two
restricted spot females (25, 3) sired one self, three re-
stricted, and one moderate. This is in line with expecta-
tion if spotting is dominant, the moderate in this case
possibly representing the homozygote.

The same restricted spot male (8) crossed (32) to a
solid-white (22) sired two solid-white and two completely
self. This is in line with the assumption that the white
female was homozygous for self, ss, and heterozygous
for color, Ww, or that spotting and white are both allelo-
morphic with self and that she was carrying self. The
male would then be wtw, w' standing for restricted or
moderate spotting.

Crosses involving greater amounts of spotting may
now be considered.

The self male (21) crossed (52) to a considerable spot
female (31) sired one self and two considerable. When
crossed (58) to a moderate spot (30) he sired three self
and two considerable. These results show that consid-
erable segregates from self and that a greater degree of
spotting may be produced from a less by crossing to self.
Modifiers are indicated.

The same male was also crossed (57) to a considerable
(34) and sired two restricted. In this case modifiers
may have been assorted to produce restriction, but the
female (34) was derived from a cross (34) of a consid-
erable (24) by a restricted (3) each of which was known
to carry self. She may therefore have been of composi-
tion w"w'; w”, much spotting, being derived from her
considerable parent and w', little spotting, from her re-

No. 629] INHERITANCE OF WHITE SPOTTING 479

stricted parent. She could then produce restricted off-
spring, w'w, when crossed to self.

Offspring of self by spotted known to carry self are’
therefore twenty-one self and twenty spotted.

‘The considerable spot male (24) was crossed to various
spotted females known by these or other crosses to pro-
duce self. With a restricted (3) he sired (34, 49) two
self, one restricted, three moderate, and three consider-
able. With a moderate (32) he sired (53) two self, one
restricted, one moderate and one extreme. With a re-
stricted (28) he sired (44) one self, one restricted, one
moderate, and one considerable. With a moderate (30)
he sired (45) two considerable and oné extreme. With a-
restricted (13) he sired (47) two self, two considerable
(one of which graded toward extreme), and one very ex-
treme. This last cross is interesting for the offspring
vary far in both directions from the parental types.

Crosses of spotted by spotted when both carry self
have produced twenty-three spotted to eight self which
is very close to the three-to-one expectation.

The considerable spot male (24) above mentioned when
crossed (41, 46, 50, 51) to his considerable spot mother
(23) sired five considerable and two restricted, the segre- `
gation being striking through failure of any moderates
to appear. This is in line with the supposition that the
mother (23) was carrying little spotting and was there-
fore of composition w”wt. The cross might therefore be
wrw X ww = 5(w"w"™, ww or ww!) + 2(wiw).

The same male (24) (w™w) was crossed (48) to a solid-
white half sister (29) from the same mother (23) (w™w')
by a white male (W?). There were produced two ex-
treme spot. The white female’ (29) may therefore have
been Ww” and the extreme spot offspring ww”.

The failure of anything higher than restricted spot-
ting to occur among the offspring of restricted by self,
although cats with considerable may carry self, indicates
that there may be allelomorphic factors determining
different degrees of spotting. In any case it appears

480 THE AMERICAN NATURALIST [Von. LIII

that self is recessive to spotting and that color is re-
cessive to solid-white. The principle is suggested that
there is a quadruple allelomorphic series:—W, solid-
white; w”, much spotted; wt, little spotted; and w, self,
with dominance in the degree of decreasing pigmenta-
tion. Crosses of white to self and of spotted to self would
be of value in checking this principle. Any one white cat
might throw besides white either much spotted, little
spotted, or self; a much spotted might throw besides
much, either little or self, and little should throw little
or little and self. If three distinct types were produced
from any one white or spotted cat crossed to numerous
self cats, this would demonstrate modifiers of consider-
able importance or disprove the hypothesis of allelomor-
phism suggested.

Attention should be called to an interesting but unex-
` plained relation that exists between yellow- and white-
spotting. ‘‘Self’’ tortoiseshells have yellow hairs closely
intermixed with non-yellow. This makes it very difficult ©
to determine degree of ticking in such animals. Tortoise-
shells with restricted white-spotting tend to have yellow
separated into patches, while further extension of white
separates yellow and non-yellow areas still more. Sep-
aration of yellow into patches appears not to be corre-
lated with amount of yellow.

GENERAL SUMMARY OF INHERITANCE oF Coat-CoLoR IN
Cats

It may be of interest to summarize very briefly the
genetic’ data thus far collected on coat-color in cats.
Ratios are not significant since fraternities from homo-
zygous dominants and heterozygotes are included
together.

Maltese dilution, m, is presumably a simple recessive
to intensity, M. Intense by intense have produced 41 in-
tense. Intense by dilute have produced 37 intense and 23
dilute. Dilute by dilute have produced 18 dilute.

Solid-white, W, evidently acts as a simple dominant

No. 629] INHERITANCE OF WHITE SPOTTING 481

overcolor,w. It is true-breeding in the hands of fanciers.
White by color (amount of white-spotting undetermined)
have produced 3 white and 4 colored (one near-self).
Table I shows summaries for white and white-spotting
of determined degree. It is obvious that although exten-

TABLE I
Offspring
P: e | -| i j
= Solid White, spotting, ele Bye | Drouin gon ienaa eal
W E ting, C Me REO S
WXXW: oon 3 1
WO: RA 1 2
Wer, Peer 7i 2
Da SIU eC 5 2
CXM o o 2 2 1 1 2
GKE- on 1 6 4 2 5
RXR = e. 1 3 1
Spotted Xspotted 3 13 6 8 8
CXS. eee 2 2 1
EN S 2 3
E a A E 16 17
Spotted X sell.. 4 18 21
poe ere ates, 13

sively pigmented animals appear among the offspring of
cats showing much white there is little tendency for a
kitten to show more white than appears in either parent.

Table II gives a summary of the results thus far col-

TABLE II
| Oftspring
Parents | Black Yellow Tortoiseshell
rete |e le te 9
Yellow & X black? nn tee ores fii B I 5l
| 164 if
eC ee 67 13 1 6 :
re = X yellow 2 a ee | 20 1

ee a E

a g’ X tortoiseshell 2 rgd 45
s 3 3
EEE E T E awe 48 S: 61 57 49
aS E X tortoiseshell 2 Ero 321 18) 37 t D
3 yi 4
E T 5 | 18| 44 ie
Yellow d X yellow Ẹ a 48 | 40 3 from one pair
Yellow € X yellow Ẹ no. 23........ 3 1 4 1

Back d X yellow 9 no. Hog Cu es 2 1 2 2

482 THE AMERICAN NATURALIST [Vou. LIII

lected in reference to the inheritance of yellow. Don-
caster’s? summaries from fancy breeders and from
Little’s data are given, kittens of undetermined sex being
omitted. The three tortoiseshell females from one pair
of Doncaster’s yellow by yellow may be readily explained
if it be supposed that the mother was an, extreme yellow
variant of the heterozygote Yy, comparable with my
cream female number 23. Anomalous black females may
be similarly heterozygous. Anomalous blacks and tor-
toiseshells are to be expected from anomalous yellow
females. Anomalous offspring are recorded in italics in
Table II. ?

. As regards banding, certain creams and blacks could
- not be classified and are consequently omitted from the
summaries. Lined by lined have given 2 lined and 2
blotched. Lined by striped have given 17 lined and 2
striped. Lined by blotched have given 12 lined and 4
blotched. Striped by blotched have given 19 striped and
8 blotched. Blotched by blotched have given 4 blotched.

As regards ticking, it is necessary to omit all yellows
and many tortoiseshells, as well as some with much white.
Hxtremely-ticked by little-ticked have given 4 extremely-
ticked. Extremely-ticked by black have given 4 ex-
tremely-ticked. Much-ticked by little-ticked have given
1 much-ticked and 3 little-ticked. Much-ticked by black
have given 2 much-ticked and 3 little-ticked. Little-ticked
by little-ticked have given 5 little-ticked and 1 black.
Little-ticked by black have given 7 little-ticked.

2 Doncaster, L., ‘‘ On Sex-limited Inheritance in Cats, and Its Bearing on

the Sex-limited Transmission of Certain Human Abnormalities,’’ Journal
of Genetics, June, 1913.

SOME HABITAT RESPONSES OF THE LARGE
WATER-STRIDER, GERRIS REMIGIS
SAY.

C. F. CURTIS RILEY

THe New YORK STATE COLLEGE OF FORESTRY AT
Syracuse University, SYRACUSE, NEw YORK

IV. DESCRIPTION OF AND EXPERIMENTS IN CONNECTION
WITH Brook HABITAT AT SYRACUSE

1. Description of Habitat.—Some further experimen-
tal work, much like that which previously has been con-
sidered, was done near a small, rapid stream (Figs. 4, 5),
approximately 4.5 miles southwest of Syracuse, New
York, in the late summer of 1918. The stream flows in an
easterly direction, into Onondaga Creek, its source being
a spring in the hills, forming the western side of Onon-
daga Valley. Water-striders, Gerris remigis, are com-
mon in certain situations on the surface of this brook
(Figs. 4, 5). The water in the stream is clear, and its
channel contains silt, gravel, small and large rocks.
There is little rooted aquatic vegetation growing along
the greater part of its course. At certain places the
current is quite rapid, but even at those points where it
is swiftest, there are small areas of quieter water, pro-
tected by rocks, or points of land jutting out into the
stream. In such situations (Fig. 4) are found water-
striders, singly and in small groups of two, three, or
four individuals. Occasionally there is a short reach of
quieter water, sometimes protected by trees (Fig. 5), on
the surface of which water-striders are found in small
numbers. Near the headwaters there is a large pool,
.formed mainly by an artificial dam, and this is separated
into two parts by the decaying trunk of a large fallen
tree. Chara grows rankly and in great mats in the pool.
On its surface water-striders live in large numbers (Fig.

483

484 THE AMERICAN NATURALIST (Von, LII

6). At this place, I have captured gerrids by the hun-
dreds. I have examined this pool and the immediate
vicinity for two successive seasons and I am convinced
that they breed here from year to year. They undoubt-
edly hibernate, in large numbers, along the shores of

Fic. 4. Detail of small, rapid brook near ieee showing spring conditions
(May). Arrow indicates direction of current. a, areas of quieter water, formed
by points of land, jutting out into the stream Fater it riders, steti remigis,
are found, in small numbers, in such situations, (Original. Whitney.)

this pool. In fact I have found a few of them hibernat-
ing in interstices where the shore slightly overhangs
the water, and also among dead leaves and other vegeta-
tion at points from a few inches to three yards away
from the pool (Fig. 6).

2. Methods.—The experiments were performed on the
shore of the pool, which is a large one for such a smali
brook. The dimensions of this body of water are ap-
proximately 55 X 17 X 2.5 feet. The shore, where the
experimental work was done, is flat and its surface is only
a few inches higher than that of the water (Fig.6). Back
of this flat area, a little more than three yards away from
the pool, there is a hill with a moderate slope. Only
those experiments will be considered here that were car-
ried far enough to evince fairly definite results. The
water-striders ‘used in these experiments were taken
directly from the surface of the pool. Different indi-
viduals were used in each experiment. All these experi-

No. 629] HABITAT RESPONSES OF WATER-STRIDER 485

ments were carried on at Syracuse in the afternoon, at
which time there was considerable reflection of the rays
of the sun from the surface of the water.

3. Responses When Facing Away from Pool.—The first
set of experiments deals with gerrids that were placed
on the ground one yard away from the margin of the

Fig. 5. Detail of small, rapid brook near Syracuse, showing spring conditions
(May). a, reach of quieter water, protected by trees; water-striders, Gerris
remigis, are found in small numbers in such situations. (Original. Whitney.)

water. The heads of the gerrids were pointed directly
away from the pool.’ In general, the responses of the
water-striders, in those experiments performed under
like conditions, showed much similarity. Therefore, I
give a condensed account of several typical experiments.

Experiment VI.—The water-strider is placed on the ground. It imme-
diately turns and faces the pool, and at once begins to jump toward the
water. While the gerrid does not turn away from the pool, it jumps
toward it slowly. Thirty seconds after being placed on the ground, the
insect is back on the surface of the pool.

Experiment V1I.—After being placed on the ground, the gerrid turns
toward the pool, and jumps in the direction of the water for a distance
of one foot. During one of the jumps, the body is so oriented that the
head is turned away from the pool. The water-strider continues to
jump, but is now moving away from the water. It proceeds in the
same direction until it reaches a point four yards away from the pool

486 THE AMERICAN NATURALIST [Vot LIN

and part way up the side of the slope. This gerrid is observed for ten
minutes and is still jumping away from the water.

Experiment VIII.—This water-strider jumps away from the water for
a distance of five inches. It now turns directly to the right, which posi-
tion places the long axis of the body parallel with the margin of the
pool. I am not sure whether this turn is due to the body striking
against some object during a jump, or whether the water-strider makes
the turn as a result of some other stimulus. The gerrid jumps and
walks parallel with the margin of the pool for a distance of two yards.
In a few seconds it turns a little more to the right, so that it is now

Fic. 6. Detail of large pool at headwaters of small, rapid brook near Syra-
B showing spring ey re (May). Chara grows in great abundance in this

a, surface of on which water-striders, Gerris re Stipa are found in
targe numbers; b, S ka and dead leaves among which water-striders hiber-
iate; c, overhanging shore of pool in interstices of which peiceetnnipeate hiber-

o
aS 4, shore of pool where experiments were performed; e, artificial dam built
of ESR (Original. Whitney.)

jumping obliquely toward the water. The insect continues to move in
the same direction until it reaches the pool. This gerrid is back on the
surface-film of the water in twenty-five seconds after the experiment
began.

Experiment IX.—The water-strider is placed on the ground and it
immediately turns to the left, thus placing the long axis of the body
parallel with the margin of the pool. The creature jumps straight ahead
for a few inches, and then turns obliquely toward the water. It moves
along rapidly and soon attains the surface-film. This gerrid reaches
the pool in twenty-five seconds from the time it first was placed on the
ground.

Experiment X.—This water-strider at once turns in such a manner

No. 629] HABITAT RESPONSES OF WATER-STRIDER 487

that the longitudinal axis of the body is placed parallel with the margin
of the pool. The gerrid walks for five inches in a path parallel with the
shore of the pool. It now turns so that its head points away from the
water, and jumps in that direction for two feet. The insect again
turns, this time the head being directed toward the water, but it walks
in that direction for two inches only. It turns to the left, thus again
placing the body parallel with the margin of the pool. The gerrid
jumps straight ahead for four feet, and again turns directly toward tlie
water, jumping rapidly along a practically straight path until it reaches
the pool. e creature is back again on the surface of the pool in forty
seconds after the experiment began. This water-strider was active from
the time it was placed on the ground until it reached the water.

Experiment XI.—The gerrid walks away from the water for a distance
of two inches. -It then turns to the left and jumps along a path parallel
with the margin of the water for four inches. The insect now makes a
turn of ninety degrees, so that its head is directed toward the pool, and
jumps rapidly to the water. In fourteen seconds after being placed
on the ground, the gerrid is striding back and forth on the surface of
ihe pool.

In these experiments I have condensed the statements
in such a manner as to give prominence to the factors. of
time and direction. With reference to these particular
elements, the experiments are typical of many others
recorded in my field notes. It is evident that a large
majority of the gerrids get back safely to the water, only
one out of the six failing to do so. This was the indi-
vidual used in Experiment VII. A large number of ex-
periments furnish similar results. Although there were
a number of random and trial movements, the water-
striders returned to the pool with a fair degree of prompt-
ness. The total amount of time consumed by all the
gerrids was 12 minutes and 14 seconds. The average
time taken to reach the water was 2 minutes and 24 sec-
onds. Omitting Experiment VII, the total amount of
time necessary for all the gerrids to get back into the
pool was 2 minutes and 14 seconds. With this experi-
ment omitted, the average time consumed in reaching the
water was 264 seconds. These results are indicated in
Table I. The results of two other sets of experiments
are indicated in Table II and Table III. Attention is
directed to the similarity in the records of the three series
of experiments as expressed in the tables.

488 THE AMERICAN NATURALIST [ Vou. LIII

TABLE I
TIME CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE

Experiments

Heads Directed Away from Water | Time Consumed | Responses
Number of Experiment | Minutes Seconds | Successes, Failures
a A E A E A., 30 | +
ME Src co ie ee are er E L 0 os |
by ia ERATA A ia E E AE 0 25 + |
Z.. 0 25 y opati
RA a Ea S EA E ois 0 14 To
Totals 6 . bie teteeeeseeteeneneecey 12 ee i | 1
Averages. . aes 2 2} | o | o}
Totals, o mitting experiment t VII.. ae ts | 3 l4 | 5 `
Averages, omitting experiment VII..... | 0 264 | 1 0

TABLE II
Time CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE
1 YARD

Experiments
Heads Directed Away from Water | Time Consumed Responses ie
Number of Experiment | Minutes | Seconds Successes | Failures
et T ee Oe ee + ie
DO SD. GUN ea es SOON ctr as a a rT | 0 | 20 A a
OD ER a ed a aye nasa Te, Sie eee. P .
sman EE ei E98 La .
p PA S ER ETA SEM A saa EN EA it te ! 0 | 35 aS .
POR o ene a ae a et | 1 | 0 i ->
PRIORI CIE E EE | 3 | 6 6
AVeTageS. -onoo ...s.s--+----+++| 0 o o TEE

|
|

4. Responses When Facing Pool.—A condensed state-
ment will be given now of experiments in which the
heads of the water-striders were turned toward the pool.
As before, the gerrids were placed one yard away from
the water. Special attention again was directed to the
factors of time and direction.

Experiment XXX.—The water-strider jumps three inches directly
toward the water. It then turns to the right, jumping parallel with
the margin of the pool for one foot. The gerrid again turns slightly to
the right, being now in a position oblique to the pool, and continues
jumping away from the water for two feet. The insect turns to the left,
so that its head points obliquely toward the pool, and jumps in a

No. 629] HABITAT RESPONSES OF WATER-STRIDER 489

TABLE III
TIME CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE
OF 1 YARD
Experiments
Heads Directed Away from Water Time Consumed Responses
Number of Experiment Minutes | Seconds | Successes | Failures
s a a e A ae 0 20 +
De, Ai’ A O E E A E AA E 0 -f |
E aa CAD EN e A A E E ee 15 25 Z —-
te VL ris. Gas kaa See 0 +
ASVI oe a ee 0 50 + |
p a a r enoa E E E S E T a OF). as Pe
eaea 6. Se eee are eas ie EOE oh as eae
Ave 4 2 | 5543s of | 0}
Totals, pes A experiment XXVI.. re | 2 30 5 0
IAN verages, omitting experiment XXVI.. r ae ae 1 0

straight path until it reaches the water. This water-strider consumed
twenty-five seconds in making the journey to the pool.

Experiment XXXI.—This gerrid jumps in a zigzag course toward
the water, arriving on the surface-film of the pool in ten seeonds from.
the time it was first placed on the ground.

Experiment XXXII—The water-strider moves toward the pool,
jumping in a direction slightly oblique to its margin and gaining the
water-film in twelve seconds.

Experiment XX XIII.—The path taken at first, by this water-strider,
is toward the pool, but after jumping for a distance of four inches in
that direction, it turns obliquely to the right, still jumping toward the
water. The gerrid is back on the surface of the pool in eleven seconds
from the time it was placed on the ground.

Experiment XXXIV.—This hemipteron takes a position so that the
body is slightly oblique with reference to the margin of the pool. The
gerrid jumps along a straight path toward the water for two feet. Ti
now turns so that the long axis of the body is parallel with the margin
of the pool. It jumps straight ahead for one yard, when it turns toward
the water, arriving at the pool in ninety seconds.

Experiment XXXV.—tThis gerrid turns to the right, as soon as it is
placed on the ground, and jumps for a distance of two feet in a direc-
tion parallel with the margin of the pool. It then makes a turn of
ninety degrees to the left, thus pointing its head directly toward the `
water. The creature jumps in this direction until it reaches the pool, -
twelve seconds after the experiment began.

The results evinced in these experiments are typical
of the results obtained in many others not recorded
here, except a few which are indicated by tables. It is

noticed that the water-striders reached the pool much

490 THE AMERICAN NATURALIST [Von. LIII

more promptly than was the case when they were placed
on the ground with their heads directed away from
the water. All the gerrids reached the water—as was
generally the case in many other experiments of a simi-
lar character—with but a limited number of random
movements. The only prominent exception to this was
the gerrid used in Experiment XXXIV. Usually, there
were one or two individuals that displayed this lack of
promptness. All the gerrids employed in the six experi-
ments consumed a total amount of time of 2 minutes
and 40 seconds. The average amount of time necessary
to return to the pool was 26% seconds. If Experiment
XXXIV should be omitted, it is evident that the total.
amount of time consumed by five water-striders in reach-
ing the pool was 1 minute and 10 seconds. The omission
of this experiment reduces the average time, consumed in
reaching the water, to fourteen seconds. These results

TABLE IV

TIME CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE
OF 1 YARD

Experiments
Heads Directed Toward Water | Time Consumed | Responses
Number of Experiment | Minutes | Seconds | Successes | Failures
| | RRA EEE on Soar
Mei, co eee S tee 25 bo
Eaa E A E E a 0 10 Ao o
OMEN SS ie ca lee N e aao 0 12 Hop
SRAL, eee kas ol ieee h 0 11 + E
XXXIV. 0 90 + E
Bio. ig E E S PO E e an aa 0 12 + ee
Spern w Cues Sees A E pa Wk wala 2 6 es
Site we OR e se E we wee ek eed 0 264 1
Totals, pA ERREAREN t XXXIV. 1 10 gx
Averages, omitting experiment XXXIV. 0 14 ee ees

are shown in Table IV. The results of other experi-
-ments of a similar character are indicated in Table Y
and Table VI.

5. Responses When Parallel With Pori — Some experi-
ments were performed with water-striders having the
long axis of the body parallel with the margin of the
pool. In all other respects, the conditions were similar

No. 629] HABITAT RESPONSES OF WATER-STRIDER 491

TABLE V
TIME CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE

Tao
Heads Directed Toward Water | Time Consumed | Responses

Number of Experiment | Minutes | Seconds | Successes | Failures
V : i -| Ot a0 ie
v ct 15 =
XXXVIII e -| 0 17 +
EE, UP. Gee PINE aren fete ee mh ene tee | 0 22 +
AN se 0 Ee +
De titi re a ce leo ees | +
$A ho os ae Vegi ee eo 52 6
Katka E S seen aaa E | 184 1

TABLE VI

TIME CONSUMED BY WATER-STRIDERS IN REACHING WATER FROM DISTANCE

Experiments
Heads Directed Toward Water | Time Consumed Responses
Number of Experiment . | Minutes Seconds Successes Failures
s K ee o 13 +
LI. 0 i +
DEE iae eas e o eee 0 12 +
BY oe a a 0 14 eje
PAV e a a a a 0 12 ES
LVI. | 0 16 +
Totals 6... A l 24 6
Averages.... o 14 1

to those when the heads were directed toward and when
they were directed away from the pool. The results were
much like those evinced in Table II, except that the time
consumed in reaching the water was slightly greater in
the majority of cases. There was a little less prompt-
ness, perhaps, in moving toward the water and a greater
number of trial directions. Occasionally a gerrid did not
reach the pool at all.

6. Responses When Not Oriented with Reference to
Pool.—A number of other simple experiments were
carried out near the large pool in the brook previously
mentioned (Fig. 6). In these the water-striders were
not oriented with reference to the position of the pool at
the beginning of each experiment. Forty gerrids just

492 THE AMERICAN NATURALIST [ Von. LIII

captured from the surface-film were put into a small
wooden box. This was taken to the place where the
other experiments were performed (Fig. 6). It was then
inverted and all the water-striders carefully shaken out
on to the ground one yard away from the water. It was
of course impossible to watch in detail every. gerrid, but
it was possible to observe how many of the hemipterons
reached the water. The majority of them were back on
the surface-film within fifteen seconds after being placed
on the ground. All but two individuals had reached the
water within thirty-five seconds after the experiment
began. At the end of one minute of time all the gerrids
were on the surface of the pool. Sometimes a water-
strider was not successful in reaching the pool. These
statements are fairly typical of the results of many other
similar experiments.

A series of experiments of a similar character was
undertaken in which the gerrids were placed on the
ground three yards away from the pool. As in the ex-
periments one yard away from the water, the hemipterons
found the surface of the pool with reasonable promptness
and directness. The greater number reached the water
within forty seconds from the time that they touched the
surface of the ground. In the majority of these experi-
ments, all the water-striders were back on the surface of
the pool, 2 minutes and 5 seconds later. In each of two
different trials, out of a total of six, there were two
gerrids that jumped away from the pool and had not
reached the water at the time my observations were dis-
continued. I believe that vision was the chief factor em-
ployed in directing the gerrids to the water in the
experiments when forty individuals were used at each
trial.

T have not yet observed the results of placing the water-
striders on the ground in large numbers farther away
from the pool than three yards. Nor have I made any
trials, either in the vicinity of Urbana or Syracuse, with
the gerrids for a greater distance from the water than
four yards.

No. 629] HABITAT RESPONSES OF WATER-STRIDER 493

V. DISCUSSION or OBSERVATIONS AT WHITE HEATH

1. Initial Locomotor Responses.—It is an interesting
fact that, just previous to the drying up of the pool, in
which the water-striders were living, there were no re-
sponses on the part of the gerrids which indicated any
attempt to escape from the unfavorable surroundings.
Not until the water had disappeared entirely was there
any tendency to leave the place. Soon after it became
dry the water-striders began to move away from the site
of the former pool. What the immediate stimulus was, it
is difficult to say. A change in the physiological condi-
tion of the body, which might have been induced by the
drying up of the pool, would be sufficient to account for
the locomotor responses. Whatever the stimulus was, the
gerrids began to walk and jump away in different direc-
tions. But as Jennings (1906, pp. 284, 285) has pointed
out:

. movement in a certain direction is due only to the release
of inhibition. The organism moves in the given direction because it is
moving from internal impulse, and because movement in this direction is
not prevented. This possibility must be considered in all eases.

Therefore, it is not always necessary to assume that
movement is due to some very recent external stimulation.
‘Whatever the explanation may be, the water-striders
moved off in the direction in which their heads were
pointed. They continued along the same line of progress
until they arrived at some obstacle in their pathway.
Such an obstacle might be a lump of dried mud, a stone,
or a piece of driftwood. Then they usually turned to the
right or left, as the case might be, thus being deflected
from their former direction of movement. They con-
tinued along the new path until they were deflected again,
in a new direction.

2. Role of Trial and Error.—Such responses as previ-
ously have been described occurred again and again.
The various objects in the path of the water-striders
served as stimuli to turn the gerrids aside and swerve
them in another direction. First they tried one line of

494 THE AMERICAN NATURALIST (Von. LILI

progression and then they tried another. As Holmes
(1916, pp. 157, 158) has said:

Where there is “ error,” the organism tries again, and keeps on doing

so until it attains ultimate success.
This statement does not mean that all achieve success,
nor does it necessarily mean that the organism possesses
any conscious appreciation of means to an end. Cer-
tainly, I do not consider that water-striders have such an
appreciation. Frequently, on coming in contact with
such obstacles, as have been mentioned, the gerrids came
to rest with the side or sides of the body closely applied
to the object. This was due to their thigmotactie procliv- °
ities. They remained in such positions for varying
lengths of time and then moved forward again, but
- usually the direction of progression was changed. Occa-
sionally, they remained motionless in such situations
until the time set for me to discontinue.my field observa-
tions for that particular day. Sometimes individuals
crawled under lumps of dried mud, under pieces of drift-
wood, or among dead leaves. On a few occasions, a few
gerrids jumped into large cracks in the baked mud of the
stream bed. Water-striders getting into such places,
occasionally remained there, but I never have been able
to find them the day following the observation.

It already has been stated that some of the gerrids
reached the larger pool of water some distance down-
stream, and attention also has been directed to the fact
that on several other occasions, when water-striders had
been trapped in stream pools, some of their number were
successful in reaching other bodies of water in the imme-
diate vicinity. I have not observed that gerrids ever
were succesful in finding another body of water that was
situated farther away than fourteen yards. In none of
these cases that have come under my observation, have I
been able to see that there was any definite response, on
the part of the gerrids, to another body of water per se.
In many instances, the locomotor movements of the water-
striders, in so far as their final goal was concerned, have

No. 629] HABITAT RESPONSES OF WATER-STRIDER 495

proved to be lacking in definiteness, precision, and in
direction of response. Their locomotor movements were
very awkward and they stumbled along the route in a
very blundering fashion. Their method, if it can be
called such, of reaching the water seemed to be entirely
one of chance. They might blunder on to a pool of water
in the vicinity or they might not. They frequently took
the wrong direction and made many mistakes. A better
way, perhaps, to express my thought, is to state that these
gerrids pass from the site of a former pool to another
body of water by a blundering method of trial and error.
As Holmes (1916, p. 158) well has said:

The method is round about and expensive, but it is better than
nothing. It is Nature’s way of blundering into success.

It is not improbable that the method of trial and error
forms a large part of the habitat responses of arthropods.
It is certainly true that a number of writers have been
impressed with the prevalence of behavior of such a
character among the members of this group.. Among
others, this is evident from the work of Bohn (1903) in
connection with hermit crabs. Holmes (1905, p. 106) in
describing the behavior of the blow-fly larva, with refer-
ence to light, makes the following statement:

It may be said to be a form of the trial and error method minus the
element of learning by experience.

Writing of the trial and error method in the conduct
of lower animals, Holmes (1905, p. 108) states that:

The lives of most insects, crustaceans, . . . and hosts of lower inver-

tebrate forms, . . . show an amount of busy exploration that in many
eases far exceeds that made by any higher animal.
In this connection the following general statements are
of great interest, as they show the importance that is
now attached to such a method of conduct among inverte-
brates: Holmes (1905, pp. 107, 108) points out that:

The rôle played by the trial and error method in the behavior of the
lower organisms has, as yet, elicited but little comment, owing probably
to the fact that attention has been centered more upon other features
of their behavior. It may have been considered by some investigators as

496 THE AMERICAN NATURALIST [Von. LII

too obvious for remark since any one who attentively observes the con-
duct of almost any of the lower animals for ten minutes can scarcely
fail to see the method exemplified.

Jennings (1906, pp. 246, 247), also, directs attention to
this form of behavior in the following words:

In most if not all other invertebrates there occur many “ trial move-
ments” similar to those already deseribed. In many recent accounts of
the behavior of other invertebrates little mention, it is true, will be
found of such movements. This is apparently because attention has
been directed by current theories to other features of the behavior, and
the trial movements have been considered of no consequence. Often an
attentive reading of papers on “ tropisms,” ete., will reveal parenthetical
mention of various “ disordered” movements, turnings to one side and
the other, and other irregularities, which disturb the even tenor of the
“tropism,” and are looked upon for some reason as without significance
and not requiring explanation. Further, one often finds in such papers
accounts of movements which are clearly of the “trial” character, yet
are not recognized as such by the author, on the watch only for “ tro-
pisms.” In the earlier literature of animal behavior, before the preva-
lence of the recent hard-and-fast theories, one finds the trial movements
fully recognized and described in detail... .

Unprejudiced observation of most invertebrates will show that they
perform many movements which have no fixed relation to sources of
external stimuli, but o do serve to test the surroundings and thus
to guide the animal. s Holmes (1905) has recently pointed out,
in a most excellent paper, a is really a matter of common observation
on all sorts of animals. The fact that such movements are not empha-
sized by writers on animal behavior is evidently due to their being con-
sidered without significance.

In a number of recent papers the importance of trial movements in
behavior has been more explicitly recognized. . . .

I have made a statement about a final goal, but I do not
intend to convey the idea, in any way, that these insects
are endowed with even the smallest amount of prevision,
nor do I wish to be understood as assuming that because
of certain perception, on the part of the water-striders, of
the exigencies of the case, they therefore responded with
a special form of behavior suitable to meet the difficulties
of the situation. But, on the other hand, I wish to present
_ the thought that these gerrids, in moving away from their
former haunts, may or may not come upon another body
of water, if there is one in the vicinity, and that this hap-

No. 629] HABITAT RESPONSES OF WATER-STRIDER 497

pens not because of any direct or definite response or re-
sponses to the body of water per se, but rather is due
more to the fact that many of their locomotor responses
are spontaneous ones, modified frequently as to direction
and speed, mainly, by contact stimulus, many of these
movements probably being due not to some very recent
stimulus or stimuli which have any direct relation to the
body of water, but that they, more probably, are due, as
Jennings (1906, p. 285) suggests,

to the simple outflow of the stored-up energy of the organism through >
the channels provided by its structure.

3. Role of Moisture Undoubtedly it is true that water-
striders, Gerris remigis, are sensitive and responsive to
moisture. The fact that the greater portion of their lives
is passed on the surface-film of brooks and streams would
seem to be sufficiently indicative of this. Then, also, the
ability to find their way back to the stream in the spring,
having left it in the fall, frequently from distances of
three and four yards, and sometimes from greater dis-
tances, after passing several months in hibernation, is
further indication that they are sensitive to some stimulus
or stimuli, the response to which results in bringing them
back to the water.

That the migration of those gerrids from the site of a
former pool to another body of water is mainly an expres-
sion of hydrotropism, according to the manner in which
that form of response is usually interpreted, I believe to
be extremely doubtful. However, it is not my intention
to assért that moisture does not play an important réle in
the economy of these water-striders. But I do not be-
lieve that the movements of the gerrids in the dry bed of
the brook afford any definite indication that they are
direct responses to moisture. It is very improbable that,
during severe droughts and high temperatures, moisture,
- diffusing through the atmosphere, from such compara-
tively small bodies of water (dimensions 3 yds. X 2 yds.
X 5 in. and in several instances smaller than this) as
already have been indicated, impinged on the bodies of

498 THE AMERICAN NATURALIST [Von. LIII

the hemipterons in any manner that would be effective in
producing definite responses to the source of this mois-
ture, as for example positive responses, resulting in the
water-striders wandering toward the pool. This is the
more improbable when it is recalled that the gerrids were
ten yards away from the water, and in other instances,
not recorded in detail in this paper, they were even far-
ther away than this, eleven, twelve and fourteen yards
distant. I also have observed their responses in the dry
bed of a stream, when there were mon of water at a less
distance than ten yards apart.,

In this connection it may be of interest to quote a state-
ment from Weiss (1914, p. 33):

Wingless forms of Gerris marginatus, which is quite common through-
out New Jersey, when removed from a pond containing some three
thousand square feet of water and liberated at distances of one, two,
three, four, five, six, seven, eight, and nine yards from the water, imme-
diately made their way back to the water without hesitancy. Of course
their movements, which consisted of a series of jumps, were more or less
clumsy, but all n in the right direction even though purposely
headed the wron

When liberated a a distaneh of ten yards, they had some slight trouble
in getting their bearings, but after making several false starts, finally
wound up by going in the direction of the water. At a distance of
fifteen yards, a longer time and more moving around were required
before the right direction was located. At thirty and forty yards away,
they seemed to lose their bearings completely and moved aimlessly about
in all directions. Even at the end of an hour they were no nearer the
water.

The observations of Weiss were of responses of water-
striders under experimental conditions and not observa-
tions of their responses under the natural conditions of
their own environment undisturbed by any extraneous
stimulus, as was the case of my observations. However.
it is pertinent to direct attention to certain facts in con-
nection with his experiments. It is evident that the re-
sponses of Gerris marginatus, especially those individuals
that were placed on the ground seven, eight, and nine
yards away from the water, differ fron those of Gerris
remigis. Members of this species do not make their way
to water, from such distances, with the promptness and

No. 629] HABITAT RESPONSES OF WATER-STRIDER 499

definiteness recorded in the experiments of Weiss with
individuals of Gerris marginatus. I infer from the little
description recorded, that the responses of individuals of
the same species, when placed on the ground ten yards
away from the water, were more of the character of those
of Gerris remigis at such a distance from a pool of water.
While I have not observed gerrids of this species make
their way to a body of water quite so far away as fifteen
yards distant, as did Weiss in some of his experiments
with Gerris marginatus, yet I am not prepared to state
that they can not do so. However, if they are able to find
water at such a distance, I believe that the achievement
is one purely of chance, or the result of a blundering sort
of trial and error. On one occasion, I observed individ-
uals of Gerris remigis leave the site of a former pool in
the bed of a stream and although I watched them for an
entire afternoon, only one, out of a group of thirty, had
‘reached an isolated pool fourteen yards distant, when I
discontinued my observations at dusk. On another occa-
sion, six water-striders only, out of a group of forty indi-
viduals, were successful in finding a body of water four-
teen yards from the site of the pool in which they for-
merly had lived. I would expect, from my own observa-
tions of Gerris remigis, the responses of apterous Gerris
marginatus, at distances of thirty and forty yards, to be
much as described by Weiss, although I have recorded no
observations of the responses of gerrids at such distances
from water. j
I believe that alate individuals of Gerris marginatus,
during migration by flight, find bodies of water mainly
through the sense of vision, as is probably true in the
case of many different species of aquatic Hemiptera, a
subject to which Kirkaldy (1899, p. 110) and other writers
have directed attention. Recent work on phototaxis—
(Holmes, 1905a), (Holmes, 1907, pp. 160, 161), (Cole, 1907,
pp. 382-388), (Essenberg, 1915, p. 400), and (Riley, MS.)
—has demonstrated that many species of aquatic bugs
respond positively to light. Benacus and Belostoma re-
spond to light during migration. In the fall of 1908, at

500 ` THE AMERICAN NATURALIST [Vor. LIII

Mankato, Minnesota, a few hundred yards from a large
swamp, near the confines of the city, I observed them for
several nights, as they flew in great swarms, around the
globes of the street are lights. On the ground, within a
radius of thirty to fifty feet of certain of the lights, were
thousands of these aquatic bugs, both alive and dead. On
several occasions, it was possible, in thirty minutes of
time, to fill a half bushel measure with the insects. In
the fall of 1915, at Milwaukee, Wisconsin, in the vicinity
of Lake Park, between the Milwaukee River and the west
shore of Lake Michigan, I observed several occurrences
similar to those just described. In these instances, the
aquatic bugs were not present in quite such large num-
bers as in the former cases. The point of importance
here is, of course, the fact that members of the two
groups, Benacus and Belostoma, respond positively to
light during migration. Comstock and Comstock (1895,
p. 132) refer to somewhat similar responses. All these
facts add still more emphasis to-the probability that
alate gerrids, when migrating, locate streams and stand-
ing water by means of vision. It should be recalled that
such bodies of water are effective reflecting surfaces.
However, it is quite possible that both alate and apterous
individuals of Gerris marginatus are responsive to mois-
ture at greater distances than is the case with apterous
members of Gerris remigis. If this should prove to be
the case, it would be of assistance to the gerrids in find-
ing bodies of water. Further, it must be recalled that the
pond to which Weiss directs attention covered an area of
3,000 square feet, while the pools of water to which I refer
were very small in size. 7

It is probably true that many arthropods respond
readily to moisture. But there is not a great deal of ex-
perimental evidence recorded in the literature, treating
of the behavior of members of the group, that presents
definite information bearing on the particular’ phase of
the subject under discussion. The experimental work
that proved to be most nearly related to the form of be-

No. 629] HABITAT RESPONSES OF WATER-STRIDER 501

havior under consideration was found in a paper by
Drzewina (1908) on the hydrotropism of crabs, Carcinus
menas. Because of the character of this work, I sha
refer to it and quote from it at some length. This writer
makes a careful analysis of the responses of these crusta-
ceans to the sea. She noticed that when one of the crabs
was placed on the beach that it oriented itself and moved
toward the sea, even at a distance of 100 meters. Her
statement (1908, pp. 1009-1010) follows:

Parmi les réactions du Carcinus maenas que j’ai eu l'occasion d’étudier
a mon séjour au laboratoire maritime de Tatihou et à la station
biologique d’Areachon, une des plus frappantes est l’orientation du
Crabe | pi son habitat naturel. C’est un fait d'observation banal qu’un
Carcinus déposé sur la plage sė dirige aussitôt du côté de la mer, celle-ci
pouvant être distante de plus de 100 mètres. Il m’a paru intéressant de
déterminer les facteurs qui influencent cette orientation particulière.

Her observations seem to prove that both orientation to
and direction of movement toward the sea were not due
to responses to light, to the sight of the sea, to the wind;
or to gravity, but on the other hand were due to the mois-
ture given off by the sea. Observations were made every
day for more than a month, at different times of the day,
both in bright sunlight and also in cloudy weather. These
facts are brought out in the following quotation (1908,
p. 1010)

J’ai pu montrer que ni la lumière, ni la “vue” de la mer, ni la direc-
tion du vent n’interviennent dans ce phénomène. J’ai fait des expéri-
ences, et j’ai obtenu des résultats identiques, aux différentes heures de la
matinée et de l'après-midi, avec un soleil vif ou sous un ceil couvert;
les Carcinus dont les yeux ont été noircis ou sectionnés se comportaient
à ce point de vue comme des Crabes normaux. Comme mes expériences
ont été faites tous les jours pendant plus d’un mois, j’ai eu le vent
venant soit de la terre, soit de la mer, soufflant dans diverses directions,
vent trés fort, ou faible, ou nul, ce qui ne modifiait pas sensiblement le
sens de l’orientation des animaux; bien entendu, quand se vent était fort,
il pouvait aceélérer ou arréter les mouvements des Crabes.

ce qui concerne l’inclinaison de la plage, celle-ci exerce bien une
influence sur les mouvements du Carcinus, qui, souvent, se laisse en-
trainer par elle et suit, dans la descente, la ligne de la plus grande
pente; mais ce n’est pas elle qui le guide dans son orientation par rap-

502 THE AMERICAN NATURALIST (Von. LILI

port à la mer. Jai pu en effet montrer, en faisant marcher des Crabes
sur des pentes creusées artificiellement et diversement inclinées, que ces
animaux peuvent tout aussi bien descendre que monter les pentes dans
leur “ fuite ” vers la mer.

Après avoir éliminé successivement divers facteurs, je me suis arrêtée
à cette hypothèse: les Crabes se dirigent du côté de la mer attirés par
Vhumidité dégagée par celle-ci; il y aurait hydrotropisme.

Drzewina noticed the character of the behavior of the
crabs after a heavy rain. The peculiarity of this behavior
seemed to present additional evidence that the movements
of the crustaceans, previously mentioned were responses
to the moisture from the sea. At such a time the ground
was very moist. Therefore there was no longer a sharp
contrast between the land and the sea, with respect to the
amount of water vapor given off by each. The crabs did
not go directly toward the sea; but some of them moved
obliquely. to the right and to the left; others followed a
zigzag course, parallel to the sea; while still others climbed
a slope and proceeded in a direction opposite from the
sea. I will record these very interesting observations in
her own words (1908, p. 1010):

Plusieurs faits que j’ai observés viennent 4 l’appui de cette hypothèse.
Aprés une pluie abondante, le sol étant humide, quand on dépose les
Crabes sur da pente sableuse, ils ne se dirigent pas directement vers la
mer, comme ils le font d’habitude, mais ils vont d'une façon quelconque :
les uns obliquent à droite ou à gauche, d’autres vont en zigzaguant
parallélement 4 la mer, d’autres enfin remontent la pente, dans le sens
opposé à la mer. Il est évident que dans le cas présent, comme il n’y a
plus de contraste assez net entre la mer et la terre, celle-ci dégageant
également de la vapeur d’eau, l’orientation des Crabes se fait d'une
façon de la quelconque.

This observer found that, when a crab was placed in
front of a kind of dyke, which at low tide separated two
bodies of water, the animal did not respond by moving
toward either body of water, but, instead, it took an inter-
mediate direction, and walked toward the dyke. She
recorded these facts as follows (1908, pp. 1010, 1011) :

Voici un autre fait intéressant au point de vue de l’hydrotropisme:
Je dépose un Crabe en face d’une sorte de digue qui, 4 mer basse, sépare
deux masses d’eau s’éntendant à droite et à gauche. Le Crabe est attiré

No. 629] HABITAT RESPONSES OF WATER-STRIDER 503

à la fois par lune et par lautre; il prend une direction intermediaire et
va vers la digue au lieu d’aller vers une des bandes d’eau.

The responses of crabs living in shallow water differed
from the responses of those living in deeper water. When
the former were placed on the beach, they displayed a
very definite hydrotropism, but the latter, under similar
experimental conditions, evinced no such definiteness of
response. Drzewina considered such responses to be
adaptive in character. She seems to infer that the char-
acter of the behavior, already acquired, must be taken into
consideration in the interpretation of their present re-
sponses. These observations are described by her as
follows (1908, p. 1011) :

Quand on prend le même Crabe dans divers habitats, on s'aperçoit
que son orientation est adaptée aux conditions dans lesquelles il vit et
qu’elle correspond aux habitudes qu’il a pu acquérir dans le cours de
son développement. Les Crabes de hauts niveaux, ayant à subir de
courtes périodes de submersion alternant avee les périodes d’émersion,
C'est-à-dire de dessiccation relative, sont très sensibles aux contrastes de
Vhumidité et de la secheresse et, déposés sur la plage, manifestent un
hydrotropisme très net. Mais les Carcinus des niveaux plus bas, pris
sur fond vaseux se comportent autrement: déposés sur la plage, ils se
dispersent dans toutes les directions, devient facilement, et, surtout, se
terrent constamment; d'une manière générale ils sont lents, peu sensibles
aux contrastes de l’ombre et de la lumière.

4. Rôle of Vision.—On the several occasions that I have
observed the drying up of isolated stream pools, having
on their surfaces trapped Gerris remigis, I have watched
carefully in order to detect whether the sense of sight was
the principal factor in aiding these aquatic bugs to find
other bodies of water. The rôle played, directly, by
vision, is probably not of immediate importance during
their responses in this connection, except in those in-
stances when the ground is flat and level and the gerrids
are comparatively close to the water. There are various
obstacles that modify the possibilities of such an explana-
tion. If there are two or more bodies of water in theim-
mediate vicinity, it has been observed that the gerrids
are just as likely to move toward the farthest one, as they

504 THE AMERICAN NATURALIST (Vou. LII

are to move toward the nearest one. If vision were the
main factor in assisting the hemipterons in finding pools
of water, they would be expected to go to the nearest one
first. Another fact against the idea of vision being the
chief influence in guiding these insects to water is that the
dry channels of the streams, where I have made my ob-
servations, frequently have very rough and uneven sur-
faces, with small boulders, stones, lumps of baked mud,
pieces of driftwood, and clumps of dead leaves scattered
along them. When the small size of these insects and the
nearness of their eyes to the surface of the ground are
both taken into consideration, it becomes very evident
that the various objects that have been enumerated must
obstruct the view of the water-striders in a very serious
fashion. Then again, sometimes the nearest pool was
around a bend in the stream, away from the gerrids, thus
making it impossible to be seen by them at a distance.
With reference to the experiments of Weiss (1914, p.
33) it is probable that sight was an important factor in
directing the gerrids to the water, especially over the
shorter distances, one to six yards inclusive. On a bright,
sunny day, it is evident that the glistening and reflective
qualities of a body of water must be factors of importance
in attracting these aquatic hemipterons. It must be re-
called that the pond to which Weiss refers was a body of
water extending over an area of 3,000 square feet in ex-
tent while the pools to which I have referred were pro-
portionately insignificant in size. If there was a gradual
slope to the shore of this pond and if the ground, where
the experiments of Weiss were performed, had a smooth
surface free of obstructions to the view, all this should be
in favor of the idea that vision was the important factor
in directing the water-striders back again to the pond.
However, the local physical conditions are not described.
Certain experiments of Drzewina (1908) are, perhaps,
worthy of mention in this general connection. This
writer found, in her observations on the hydrotropism of
crabs, that these animals responded positively, and with

No. 629] HABITAT RESPONSES OF WATER-STRIDER 505

considerable precision, to the moisture given off from the
sea. However, in other experiments with crabs she con-
sidered that the past life of the crustaceans and the char-
acter of the behavior, already acquired, must be taken into
consideration, in the interpretation of their present re-
sponses. Crabs, living in deep water, among rocks cov-
ered with alge and beaten by the waves, when placed on
the sand, in the vicinity of the sea, did not evince definite
hydrotropic movements, but, on the other hand, their re-
sponses were of a very different character. The factor,
in these responses, of importance to the present discus-
sion is that of sight. Vision, apparently, played a promi-
nent rôle in determining the direction of movement of the
erabs. Drzewina (1908, p. 1011) has given a rather full
statement concerning these facts:

Les Carcinus de la zone basse de Fucus serratus, pris & une pointe
rocheuse (Gatteville), où ils vivent eramponnés parmi les rochers
couverts d’algues et battus par les flots, se comportent encore autre-
ment: lâches sur du sable, au voisinage de la mer, au lieu de descendre
vers celle-ci, ils se dirigent immédiatement, en ligne droite, vers des
rochers couverts d'algues, ces rochers pouvant etre sities a plusieurs
mètres de distance latéralement à droite, à gauche, ou à la limite d’eau,
ou même dans le sens opposé à la mer. Et ceci, quelle que soit la direc-
tion du vent et du soleil. Ces mêmes Crabes, déposés sur du sable clair,
légèrement humide, où, par places, se trouvent disséminées des taches
sombres de Fucus, se dirigent vers ces taches. Jamais je nai pu con-
stater, avec ces Crabes, d'orientation directe par rapport à la mer, mais
toujours une attraction très prononcée exercée soit par des saroe soit
par des touffes d’algues, par des surfaces d’ombre, en un m

Ces quelques faits montrent combien il est important, ae Vinter-
prétation des réactions, de tenir compte du passé de l’animal et des
“ habitudes ” que celui-ci a pu créer. Dans ’hydrotropisme du Carcinus
maenas, Vintervention des habitudes est des plus manifestes.

(To be concluded)

BEHAVIOR AND ASSIMILATION

DR. HENRY D. HOOKER, JR.

UNIVERSITY OF MISSOURI

I

In a discussion of Liebig’s law of the minimum (Hooker,
17), proof was given of the existence of an integrating
principle which, as Adams (’18, p. 481) points out, is
equivalent to Bancroft’s law, so called because Bancroft
(711) was the first to indicate the application of Le Cha-
telier’s theorem to biology. In fact, if it be admitted that
organisms are systems in equilibrium, it follows that they
obey the theorem of Le Chatelier. Bancroft’s formula-
tion of the law is ‘‘that a system tends to'change so as to
minimize an external disturbance.’’ But this statement.
is so broad that it fails to convey the full significance of
the theorem and apparently has led to some confusion.
It therefore seems advisable to give a detailed discussion
of the theorem of Le Chatelier in its application to biol-
ogy and more particularly to point out its relation to
other biological principles.

“It will be perceived,’’ says Troland (’17, p. 325),
that the demand... is not for new biological facts,
but for physico-chemical conceptions in terms of which a
chaos of biological facts, already at hand, can be ex-
plained or systematized.”’

Findlay (’04, p. 56) defines the theorem of Le Chatelier
as follows:

If a system in equilibrium is subjected to a constraint by which the
equilibrium is shifted, a reaction takes place which opposes the con-
straint, i. e., one by which its effect is partially annulled. ... In all
cases, whenever changes in the external condition of a system in equi-
librium are produced, processes also oceur within the system which tend

to counteract the effect of the external changes.

506

No. 629] BEHAVIOR AND ASSIMILATION 507

Let us consider, by way of example, a simple case of a
system of three phases, namely, ice, water and water-
vapor, in equilibrium with respect to temperature and
pressure. This system can exist only at 0° C. and at
atmospheric pressure. If heat is withdrawn from the
system and the pressure and volume are kept constant,
a part of the water freezes to ice and the temperature is
maintained by the latent heat of fusion. Since the for-
mation of ice would increase the volume and therefore
raise the pressure, a certain amount of vapor condenses
to water. If heat is added to the system, changes take
place in the reverse direction. Similar changes occur
when the pressure is altered at constant temperature. In
general (Findley, ’04, pp. 56, 57), ‘‘so long as the three
phases are present, no change in the. temperature or
pressure of the system can occur, but only changes in the
relative amounts of the phases; that is to say, the effect
on the system of change in the external conditions is op-
posed by the reactions or changes which take place within
the system (p. 60). If the specific volumes of the phases
are known and the sign of the heat effects which accom-
. pany the transformation of one phase’into the other, it is `
possible to predict (by means of the theorem of Le Cha-
telier) the changes which will be produced in the system
by alterations of the pressure and temperature. .. . It
should be noted that all three phases are involved in the
change.”

It is evident that these remarks apply in detail to the
behavior of living organisms. The system in equilibrium
is the organism. The external condition of the system is
the environment. The constraint by which the equilib-
rium is shifted is the stimulus. The reaction that op-
poses the constraint and partially annuls its effect is the
response of the irritable mechanism.

A stimulus is generally considered to be any change in
the relation between the organism and a factor of its en-
vironment, but no response occurs unless the change be
one by which the equilibrium is shifted, to effect which it

508 THE AMERICAN NATURALIST [Voit LII

must be, (a) of a magnitude sufficient to overcome the
inertia of the system, that is it must be a liminal stimulus
and (b) it must relate to a limiting factor of assimila-
tion) cf. Hooker, ’17, p. 204). Furthermore, the change
in the relation between the organism and a factor of its
environment may be produced either by a change in the
environment, i. e., it is heterotrophic, or by a change in
the organism, i. e., it is autotrophic, or by changes in
both.

The response is characterized by the facts that (a) its
nature and direction are determined by the stimulus but
(b) the energy is supplied by processes that occur within
the system; in other words the stimulus releases the
response. Thus in the ice-water-vapor system when heat
is withdrawn, this is supplied by the latent heat of fusion
of ice. In this simple system there is a direct relation
between the amount of heat withdrawn and the amount
supplied. Such a relation does not exist in the behavior
of organisms, nor should we expect it in such complicated
systems. A wealth of examples illustrating the applica-
tion of the theorem of Le Chatelier to the behavior of
plants and animals will be found in Bancroft’s article
(717). A brief discussion of the integrating character
of development, evolution and biotic succession has been
made elsewhere (Hooker, 717) and supplies illustrations
of these principles. It will be sufficient to reiterate here
that cells, organs and groups of organisms form systems
as well as the single organism. Correlations and mor-
phogenic responses are therefore conditioned in accord-
ance with the theorem of Le Chatelier, and it is unneces-
sary to postulate the existence of hypothetical inhibiting
substances to account for the normal behavior of parts.

Inasmuch as the reaction of a system is directed ac-
cording to the theorem of Le Chatelier, every system in
equilibrium is teleological. The means that produce the
reaction are directed to a definite end, to overcome the
constraint, and the reaction might be said to take place
in order that the system may be preserved. This is evi-

No. 629] BEHAVIOR AND ASSIMILATION 509

dently the source of the ‘‘purposefulness,’’ that has occa-
sioned endless biological discussion. The living organ-
ism, however, is teleological only to the same extent as
the ice-water-vapor system. :

These being the facts, the essential problem that con-
fronts us is a study of the physical and chemical aspects
of the equilibrium under the influence of a complex of
environmental factors. There must exist some degree of
correspondence between this equilibrium and the en-
vironment. Comparative morphological structure is evi-
dence of this and the same must obtain of chemical com-
position. The task before us is not so much to ascertain
the ‘‘function’’ of any particular substance or element,
as to acquire a knowledge of the equilibrium as a whole,
for it is only in relation to the other constituents that the
function of any one compound can be understood. The
need is consequently for a more detailed knowledge of
the interrelationships of the chemical constituents and
for determinations of how these relationships are altered
by changes in the organism itself. This can probably be
carried out more readily with plants, but the problem
demands more comprehensive analyses than any that
have hitherto been made. Investigations with these
points in view are now under way and the remarks in this
paper will serve as an introduction to them. A knowl-
edge of the conditions governing the change from one
phase to another within the organic system is a necessary
prerequisite to a more complete understanding of organie
equilibria. For the discussion of Le Chatelier’s theorem
has shown, that in any reaction all the phases are in-
volved, and if the changes, such as the sign of the heat
effects, etc., accompanying the transformation of one
phase into another are known, it will be possible to pre-
dict by means of the theorem of Le Chatelier the effects
that will be produced within the system by altering any
of its relations with the environment.

Adams (719, p. 74) says: ‘‘Irritability may not be
causally explained, but it seems to obey these general

510 THE AMERICAN NATURALIST [Vou. LIII

laws in the same measure as causal changes.” A com-
plete description is the only adequate explanation. Al-
though it is impossible to give a complete description of
the physical and chemical processes involved in the ir-
ritable mechanism, it should be evident that organisms
are irritable because they are systems in equilibrium and
as such obey the theorem of Le Chatelier.

II

When Adams (718, p. 474) says, referring to Bancroft’s
law: ‘‘In other words this is a perpetuating tendency, a
method of assimilation, of which reproduction may be
considered but a special phase,’’ he makes a serious
error. That some other principle is involved is hinted
at in the following passages (Adams, ’18, p. 474, 475) :

In addition to influences which interfere with systems as expressed by
Bancroft, there are those which reinforce or accelerate (tend to continue
or hasten activity) and do not change its character, but only the intensity
of the response (temperature, enzymes, repetition, ete.). By this method
also systems tend to be perpetuated and organisms in “ favorable” (non-
nord conditions, tend to continue their normal activities.

rndike in summarizing the laws of “ acquired behavior or learn-
ing” ican two laws. The first is essentially a statement of Ban-
eroft’s law of response to interference (discomfort or satisfaction) and
the second (exercise or repetition) is that of reinforcement.

But when Adams (718, p. 475) states: ‘‘This law ap-
pears to be a corollary of Bancroft’s law which is con-
cerned with interference or retardation,’’ he is dodging
the question.

The second principle which Adams calls the law of re-
inforcement is the basis of assimilation and the related
processes of growth, reproduction and inheritance. That
the assimilation of an organism may take place, three
conditions are essential: (a) available materials to con-
struct the organism; (b) a supply of energy; and (c)
the presence of the living organism. There are also cer-
tain formal conditions of temperature, pressure, etc.,
which need not concern us at present. In short, assimila-

No. 629] BEHAVIOR AND ASSIMILATION 511

tion is an autocatalytic process; because the end products
of assimilation act as catalytic agents for their own syn-
thesis. In green plants and the independent bacteria it is
also an endothermic process, because the end products of
assimilation have a higher energy content than the com-
pounds from which they are constructed. As an illustra-
tion of an autocatalytic reaction between inorganic com-
pounds, the formation of Millon’s reagent may be cited.
When mercury is dissolved in nitric acid, a certain
amount of nitrous acid is formed. Nitrous acid is a
catalytic agent for the solution of mercury in nitric acid,
and therefore for its own production. To quote from
Troland (’17, p. 337), who has discussed the theory of
autocatalysis :

e suggestion that the fundamental life-process of growth is the
expression of an autocatalytic chemical reaction has been made inde-
pendently by a number of investigators. It will be perceived that on
the basis of the foregoing theory of autocatalysis, this suggestion þe-

i nt

e growth of the crystal. The customary objection to this
comparison, viz., that a erystal grows by accretion whereas protoplasm
increases by intussusception, loses its force as soon as we regard living

matter as a complex mixture of substances suspended by colloidal sub-
division in water, since there is no evidence that the individual colloidal
particles do not grow by accretion. On the contrary, it is almost incon-
ceivable that these bodies, which are the real chemical units in proto-

to time, actually do coincide in general form with the curve characteristic
of an autocatalytie reaction.

In other words, the process of assimilation is like crys-
tallization from a supersaturated solution by seeding
with a crystal. The following quotation from Findlay
(’04, pp. 67-68) throws light on this process:

In general, then, we may say that a new phase will not necessarily be
formed immediately the system passes into such a condition that the
existence of the phase is possible; but rather, instead of the system
undergoing transformation so as to pass into the more stable condition

512 THE AMERICAN NATURALIST [Vou. LILII

under the existing pressure and temperature, this transformation will be
“ suspended ” or delayed, and the system will become metastable (that is
temporarily stable as long as it is not brought in contact with the new
phase). Only in the ease of the formation of the liquid from the solid
phase, in a one-component system, has this reluctance to form a new
phase not been observed.

To ensure the formation of the new phase, it is necessary to have that
phase present. The presence of the solid phase will prevent the super-
cooling of the liquid.

As to the amount of the new phase required to bring about the trans-
formation of the metastable phase, quantitative measurements have been
carried out only in the case of the initiation of crystallization in a super-
cooled liquid. As the result of these investigations, it was found that, in
the case of superfused salol, the very small amount of 1X 10-1 gm. of
the solid phase was sufficient to induce crystallization. ig et serene
of the supercooled liquid, however, can be initiated only by a “ nucleus’
of the same substance in the solid state; . . . it is not brought about by
the presence of any chance solid.

The following illustration is of interest in this connec-
tion. From a saturated solution of anhydrous sodium
sulphate, Na,SO,, at 24° C. different compounds may be
obtained by seeding with different crystals. If a crystal
of the heptahydrate, Na,SO,:7H.,O, is added, this salt
crystallizes out; if a crystal of the decahydrate known as
Glauber’s salt, Na,SO,-10H,O, is added, Glauber’s salt
crystallizes out. In this case both the solution and the
heptahydrate are metastable. It should be noted that the
seeding accomplishes two things: (a) it induces crystal-
lization and (b) determines the nature of the crystals.
Moreover since the crystals produced are hydrates, they
represent products of synthesis.

II

Living organisms accordingly perform two processes,
behavior and assimilation, and depending on whether the
one or the other process is considered, they present fun-
damentally different aspects. If we make a cross-section
(to borrow an expression of E. B. Holt) through the or-
ganism in one direction, it appears as a system in equi-
librium obeying the theorem of Le Chatelier. If we make

No. 629] BEHAVIOR AND ASSIMILATION 513

a cross-section at right angles, as it were, it appears as
part of an autocatalytic reaction. Behavior and assimi-
lation work in different planes. The theory of auto-
catalysis does not explain all biological enigmas as Tro-
land (717) intimates, nor does the theorem of Le Chatelier
account for assimilation.

However, these planes intersect; the two processes are
interrelated in the following four respects.

1. The factors of the environment that constitute the
external conditions of the behaving system are the pos-
sible limiting factors of assimilation.

2. Assimilation is an endothermic process that sup-
plies the energy expended by the responses of behavior.

3. Behavior overcomes the effects of the limiting fac-
tors of assimilation and so places the organism in what
is, under the circumstances, the most favorable situation
for assimilation.

4. Assimilation and growth eventually change the re-
lations between the organism and the factors of its en-
vironment, consequently producing stimuli to behavior.

Behavior is the process by which the organism is able
to cope with its environment, it renders its condition as
‘‘favorable’’ as the situation permits, but contributes
nothing to its increase. It is the progressive element to
which change and variability are due. It is the ‘‘guide
of life.’’ Assimilation is the autocatalytic process by
which the organism increases and multiplies, but which
would soon be brought to a standstill according to the
law of the minimum, were it not for behavior. It is the
conservative element that determines that like shall beget
like; it is the principle of heredity. An organism may be
defined as a system that perpetuates itself by autocataly-
sis and reacts according to the theorem of Le Chatelier.

LITERATURE CITED
Adams, C. C.
1918. — as.a Factor in ea Its Ecological Dynamics.
. Nart., Vol. 52, pp. 46
1919. open as a Factor in Evolution: Its Ecological Dynamics,
II. AMER. Nat., Vol. 53, pp. 55-78.

514 THE AMERICAN NATURALIST : {Vot LIH

Bancroft, W.
1911. A ASE Law. Jour. Amer. Chem. Soc., Vol. 33, pp. 92-120,
also Science, N. S., Vol. 33, pp. 159-179.
Findlay, A. :
1904. The Phase Rule and Its Applications. Pp. 1-313. London.
Hooker, H. D., Jr.
1917. Liebig’ s Law of the Minimum in Relation to General Biological
Problems. Science, N. S., Vol. 46, pp. 197-204.
Troland, L. T.
w Biological Enigmas and the Sans of Enzyme Action. AMER.
Nart., Vol. 51, pp. 321-350

ON THE USE OF THE SUCKING-FISH FOR
CATCHING FISH AND TURTLES: STUDIES
IN ECHENEIS OR REMORA, IT

DR. E. W. GUDGER

AMERICAN Museum oF Natrurat History, New York City

ARE THESE Accounts CREDIBLE?

Even with the cumulative evidence of all the foregoing
accounts, the matter seems almost incredible. Dr. East-
man, when working up the data for his Reversus paper, ©
wrote Dr. David Starr Jordan, and I also have written
him. Dr. Jordan, without having had laid before him the
later accounts found in this paper, doubts the Columbus
stories and kindly gives some data from his own wide
experience. He notes that the Remora rarely grows over
- sixteen inches long, and, although it holds so tightly that
it may be drawn out of water, is so small that it could
hardly be used as a hunting fish. He further adds that
Echeneis though it grows to a length of three feet doesn’t 7
“sit tight” but drops off the minute ‘‘the shark to which
it is fastened is drawn out of the water.’’ This observa-
tion of Dr. Jordan’s is in full accord, it may be noted, with
the records left by Columbus and his chroniclers that the
fish cannot stand access to the air, and hence they affirm
that, while it cannot by pulling be dislodged from its prey,
it may easily be disengaged by lifting both hunter and
hunted up into the air when the former at once drops off.

In this connection it is pertinent to give here an observa-
tion which I made at Beaufort, N. C., a number of years
ago. A shark was hooked off the laboratory wharf and
when pulled in was found to have an Echeneis attendant
about a foot long. As shark and Echeneis were both
drawn up, whenever the latter was elevated above the
water it would let go, drop down into the water, and take

515

516 THE AMERICAN NATURALIST [Von. LIII

afresh hold. This was repeated several times, but at last
the sucking-fish dropped off and swam away into deeper
water.

A: shark, which I hooked at Tortugas, had two Eche-
neises and one Remora on it. Being rapidly dragged up
on the beach, the attendants did not let go until the shark
was high and dry. One Echeneis was captured but the
other two ‘‘suckers’’ got back in the water, and in many
other cases I have been unable to bring the fish out with
the shark. Dr. Townsend, however, writes (1915) that he.
has taken many sharks while on the United States Fish-
eries Steamer Albatross, and that nearly all of them came
aboard with one or more HEcheneises attached. Thus it
would seem that in some cases the Echeneis holds fast and
that in others it lets go. The causes of these different
actions are not clear.

However, it is after all not a question of whether the
sucking-fish lets go when brought into the air, but whether
it can stand the strain of hauling in a heavy turtle or large
fish. Hence the question is one to'be settled by experi-
ment and if possible by mathematics.

Holder (1995) has been quoted as having tried the ex-
periment but without results since his’ ‘‘suckers’’ would
not leave the sides of his boat to lay hold of the turtles
and sharks. No record has been found of any similar ex-
periments. In the same paper Holder speaks of lifting a
bucket of water by a Remora which gripped the bottom
with its disk.

Dr. C. H. Townsend, in an article in the Bulletin of the
New York Zoological Society (1915), describes experi-
ments to test the holding powers of sucking-fish in the
New York Aquarium. <A two-foot specimen (size of disk
not noted) held by the tail lifted a pail half full of water—
total weight 21 pounds. A second, 26.5 inches long and
having a disk 5.5 inches in length, supported a pail and
water weighing 24.25 pounds. Had a deeper bucket been
used so that the water would not spill out, Dr. Townsend
thinks that the fish could have lifted an even greater

No. 629] STUDIES IN ECHENEIS OR REMORA 517

weight. Fig. 11, Plate III, is reproduced from the article
in question.

Dr. Townsend made another and even more pertinent
experiment which we will let him describe in his own
words:

By way of testing its fish-catching capacity, a shark-sucker sixteen
inches long was liberated in one of the tanks of the Aquarium contain-
ing fishes. It took hold at once, and by hauling on the cord fastened to
its tail a good-sized grouper was brought to the surface of the water,
although it could not be lifted out of the tank. When the fish began to
struggle the shark-sucker let go. When tried on a fifteen pound sea
turtle, the latter could easily be drawn to the surface.

There can be no doubt that with a line attached to a large remora
[Echeneis?] a much larger sea turtle could be hauled in without
diffculty.

My opportunities for making such experiments have un-
fortunately been very few. At Tortugas in 1914, I pulled
on the tail of a sucking-fish, stuck fast to the glass wall of
the aquarium, so hard that its muscles could be heard to
crack, and I had to desist for fear of pulling the tail off.
In 1913 in the Bight of Cape Lookout, N. C., we caught a
27.25 inch Echeneis having a sucking disk 6 inches long by
2.13 wide. When stuck to the wet deck, I pulled on this
fish so hard that I feared that I would tear it in two, but
it resisted all efforts to pull it off backwards—a pull of
possibly 50 pounds. On pulling upwards on it, it held fast
until the disk began to tear loose from the head.

Another fish 33 inches long, with a sucker 7.25x 2.63
inches, caught on the same day, was also experimented on
but it ‘was not in good condition and did not have the
‘‘grip”’ of its predecessor. A spring balance was attached
to its tail, and the fish was found to resist a pull of 14
pounds lengthwise and 17 at an upward angle of 45 de-
grees. The other fish would undoubtedly have shown
better results. In both the popping of the muscles was
noted. The literature is found to be filled with statements
that the fish holds tight so strongly that it will suffer
itself to be torn in two rather than let go. :

In ascertaining the ‘‘pull’’ that an Echeneis might

518 THE AMERICAN NATURALIST [Vou. LII

withstand when used as a fisherman-fish, the problem may
be attacked from two standpoints. One might first ascer-
tain how much adhesion the sucker would develop or
secondly work out the ‘‘pull’’ necessary to land a turtle
or large fish. This has never before been attempted, but
it is worth while trying.

My largest Echeneis had a sucking disk whose longest
diameter measured 7.25 X 2.63 inches. Assuming that it
was an ellipse in shape let us proceed to get its area. The
formula for the area of an ellipse whose longest diameters
are a and b is as follows: Area =r X a XbD/4. Substi-.
tuting and performing the operations indicated we get
14.98 square inches for the area. This is of course too
large, for no allowance has been made for the longitudinal
raphe or for the lamelle arranged like the slats of a
Venetian blind. This cannot be ascertained but we will
allow 1.48 square inches for this and thus reduce the avail-
able sucking area to 13.5 square inches. It is understood
that the fish adheres to its host by raising its disk through
muscular action and thus creating a partial vacuum. If
this vacuum were perfect the adhesion would be 13.5 X 14.7
pounds or 198.45 pounds, or in round numbers 198 pounds
would be the measure of the adhesion of the disk to the
host fish, shark or turtle.

But it is objected that the disk can not develop a perfect
vacuum, and as this is true our figures must be reduced.
Let the reduction be 50 per cent. and the adhesion is still
99 pounds, or if it be cut by three fourths the adhesion
will still be 49.5 pounds. The latter is probably too low
just as the first is undoubtedly too high. For one thing
there has not been and can not be figured out the ad-
ditional adhesion developed by the backwardly directed
teeth of the lamelle, which were formerly thought to do
all the holding. Assuming that the holding force is 99
pounds, it will be seen that when I pulled on this fish’s
tail with a pull estimated at 50 pounds the limits of its
adhesion had not nearly been reached. And yet at the
time it was noted that the muscles of the fish snapped and

No. 629] STUDIES IN ECHENEIS OR REMORA 519

cracked until it looked as if its tail would be pulled off.
Furthermore, scores of observers have noted that the fish
would allow itself to be pulled in two rather than let go.
In Dr. Townsend’s article, Mr. L. L. Mowbray, of the New
York Aquarium, is quoted as believing that the fish under
strain while under the weight of a considerable column of
water can not relax the grip of its disk. With a backward
pull on its tail the teeth setting in the epidermis of its host
would tend to keep the lamelle vertical, for the lamelle
ean only come to rest, relaxing the grasp of the disk, by
describing an are backward and downward. Whether or
not Mr. Mowbray’s suggestion is tenable, it is certain that
with a strong pull backward on its tail the Echeneis would
find it hard, perhaps impossible, to flatten the disk and let
go its hold.

The data offered above ought to convince any one that
Echeneis can hold on tightly enough to enable the fisher-
man to haul in the prey thus laid hold of by the fish, but
whether the fish, without being literally torn in two can
stand the strain of this hauling in is a question that must.
be met. The answer here is an unhesi‘ating ‘‘Yes’’!

Dr. Townsend pulled ‘on his 26.5-inch sucking-fish hav-
ing a 5.5-inch disk with a measured force of 24.25 pounds.
Let us see what a pull so great as this will do in landing a
large fish. Dr. Charles Frederick Holder has done more
than any man in America to popularize the taking of big
game fishes with light tackle. In his ‘‘Big Game Fishes
of the United States” (1903) to face page 80 he has a
photograph -of a 251-pound tuna (Thynnus thynnus)
caught on a 21-thread line tested to break at 42 pounds
dead pull. On page 115 he tells of the capture of a 419-
pound black sea bass (Stereolepis gigas), and to face page
116 gives a photograph of it. The size of the line is not
given but it is elsewhere specifically stated that it was not
over 24-thread having a breaking strength of 48 pounds.
In his ‘‘Big Game at Sea” (1908) on page 129 Holder
gives a table of nine tunas ranging from 94 to 251 pounds
in weight and none caught on lines of more than 24

520 THE AMERICAN NATURALIST (Vou. LIII

threads and many on 15- and 18-thread lines. On the
following page (130) he tabulates the weights of eight
black sea bass ranging from 327 to 436 pounds caught on
number 24 lines. Again on page 109 he states that Edwin
vom Hofe took a 600-pound sawfish on a 15-thread line
breaking at 30 pounds dead pull. On pages 166-167 of
his ‘‘Recreations of a Sportsman on the Pacific Coast’’
(1910) Holder describes the catching of an 8- or 9-foot
shark with a 9-thread line, and on page 169 he describes
the taking of a 12-foot, 310-pound shark with a tuna rod
and reel, the line of which was not heavier than 24
threads. Examples might be still further multiplied, but
enough have been given to establish the facts.

Now the tuna is probably the hardest fighting game fish
in the world taken on a light line, while the black sea bass
and the sharks are undoubtedly the heaviest dead weights
of any fishes taken with rod and reel. Thus the argument
is that if these fishes can be taken on lines breaking at
dead weight pulls of from 18 to 48 pounds, then an Eche-
neis, which, suspended by a cord around its tail, supports
a dead weight of 24.25 pounds,.could easily be used to
draw in fish and turtles of the sizes noted throughout this
paper. And this without any danger of tearing its tail off.

To the present writer, all the evidence at hand sustains
and confirms the stories of the living fishhook from the
time of Columbus to the present day.

Postscript

Since this paper was written I have chanced upon an
article by F. Tamborini, bearing the title ‘‘Jagd- und
Kunstschafterdienst im Meere,’’ published in ‘‘Die
Natur,’’ 1900, Vol. 49, pp. 234-235. Examination of this
shows that it contains nothing new, but note is made of it
here that this may be understood and for the sake of
completeness.

BIBLIOGRAPHY
ien Ulyssi
3. De Piseibus et de Cetis, p. 368, fig. Bononiæ,

No. 629] STUDIES IN ECHENEIS OR REMORA . 521

— y Morales, Antonio.
1883. Cuba Primitiva: Origen, Lenguas, Tradiciones de Historia de
toe Indios de las karai Mayores y las Lucayas, p. 277.
Habana, 2 ed
Banfield, E. J.
1908, sia, 1913 eds. The Confessions of a Beachcomber... in —
ila Queensland, pp. 239-246. London.
Bernaldez, n
1856. kan de los Reyes Catolicos, pp. 54-55. Granada. Madrid,
1870. MS. ante 1500.
able erm
solaio.
ee ae Ann
In the "E the Tropics, and the Roaring Forties, pp. 168-
69. New York.

\

Breton, Raymond.
1665. Dictionnaire Caraibe-Francois, p. 70. Auxerre. Réimprimé par
Jules Platzmann (edition fac-simile) ; Teipsig, 1892.
Brussel, Th. v
1729. Natuurkundige Beschryving, pp. 84-85, fig. Amsterdam,
Bullen, Frank
m "y the Deep, p. 123. Chicago.
Casas, Fray Bartolome de las
1875. Historia de las aimi, tomo II, p. 56. Madrid. (MS. about

Cobo, Barnabe.
1891. Historia del Nuevo Mundo, Vol. II, pp. 180-181. (Edited by
arcos Jiminez de la Espada). Sevilla
keni
1911. Attice on sen EDE Encyclopædia Britannica, II ed., Vol.

‚P
(Columbus, an it),
1870. Extract from Journal of Second Voyage. In Chap. 126 of
pe Bernaldez’s Historia de las Reyes EN MS. ante
1500, published at Granada 1856, and Madrid 1
Columbus, Ferdinand.
e History of the Life and Actions of Admiral Christopher Columbus
; and his Discovery of the West Indies Called the New World.
In Churchill’s Voyages, Vol. II, 1732, p. 560. Also in Kerr’s
Voyages and Travels, Vol. II, 1811, pp. 7 and 117.
Dampier, William
1729. Dampier 2 PEx yo II, part 3, A Discourse on the Winds,
; ed., 1729. (Edited by John Masefield
na TERE a Stebel London, 1906.)
Eastman, C. R
1915. Early Figures of the Remora. Nature, Vol. 95, pp. 344-345, 2
fi

gs.
1915a. Columbus on the Remora. Copeia, no. 19, pp. 11-12.

522 THE AMERICAN NATURALIST [Von LIII

1916. The Reversus. A Fishing Tale of Christopher Columbus. Sci-
entific Monthly, pp. 31-40, 5 figs.; Americana, Vol. IT, pp. 438-
446; Bulletin Pan American Union, Vol. 43, pp. 368-371.
Eden, Richard.
The Decades of the Newe Worlde or West India, ete. Wrytten
in the Latine Toungue a Peter pron of Angleria, and
racea into ec she by Rychard Eden. Londini. :
First B of the Decades of Aeey Ocean, 1511, p. 76 pet
A. D. poms 14-15 in The First Three English Books on
America (1511?) Being Chiefly Translations, Compilations,
ete., by Richard Eden. Edited by Edward Arber, London,
1895).
Galvano, saree
1601 he cept halve of the World from their first Original unto the
pes of our’ Lord 1555. Briefly Written in the Portugall
Tongue by Antonio Galvano, Governor of Ternate, the Chiefe
Island of the Malucos; corrected, quoted, and now published

in rhs Sec Peep ished 1563. = aan in Hakluyt inant
Vol. 30 n Portuguese [text of 1653] and in Hakluyt’s
Eng. ed. of 1601, edited by C. R. D. Bethune.)
Gesner, Konrad von.
1558. rhe Animalium, Book IITI, pp. 483-484, De Gviacano vel
Rev isce Indico, fig. Zurich.
1875. Das Fiahbuch, p L, fig. Zurich.
Gill, peers Wyatt
Life in ‘an Southern Isles, or Scenes and Incidents in the
South Pacifie om New Guinea, p. 292.
Gomara, Francisco Lopez
15 Historia mi te de las Indias, folio XIII. Medina del Campo.
Gosse,
1851, Matai’ History, Fishes, p. 270. London.
Haddon, A. C.
1889, _The manana of the Suckerfish (Echeneis) in Turtle-fishing.
Nature, Vol. ag
1890. The es n eae of the Western Tribes of Torres Straits.
Journal pin ni an Institute of Great Britain and Ire-
land, Vol. 19, pp. 349-350.
1890a. [Use of the eigenen va Echeneis naucrates, for Fishing in
Torres Straits. | k Lore, Vol. I, p. 80.
1901. Head Hunters, as ers and Brown, pp. 155-156. London.
1904, 1908, Reports Cambridge Anthropological Expedition to
Torres Straits: "Vol. V, 1904, pp. 44-46, 67, 92; Vol. VI, 1908,
; 41-42; Vol. IV, 1912, pp. 163-165, figs.; 221, fig.
Hernandez, Franc ‘
1665 bere Libros. De la Naturaleza y Virtudes de la Plantas y
Animales que Estan Recividos en = Uso de Medicina en la
Neuva Espana, pp. 183-184. Mexico City.

No. 629] STUDIES IN ECHENEIS OR REMORA 523
Herrara, Antonio de.
1601. Hi ae General de las Indias oe Translated by
. John Stevens, Vol. I, p. 130, 1725. Als
chore’ Voyages, Vol. V, p. 631, 1788: and in Kerr’s
yages, Vol. II, p. 322.
en, S chases Freder ick.
1899. Stories of Animal Life, pp. 42—43, fig. New Yor
1903. The Big as Fishes of the United States, pp. ora 80, 115, 125.
New Yor

1905. ker Remoras. Scientific American, Vol. 93, pp. 162-163, fig.
1906. Hal E Hone with Fishes, Reptiles and Birds, p. 80, fig. New
York

1908. Big Gam t Sea, pp. 109, 129-130, 158. New Yor
1910. Geena, of a Sportsman on the Pacifice Coast, pp. ee
New York.

Holmwood, Francis.
1883. [Fishing with the Chazo in Zanzibar.] In Fisheries of Zanzi-
ar, in Official plea International Fisheries Exhibition,
London, 1883, pp. 381-3
1884, Lot z Employment of * e “Re mora by Native Fishermen on
ast Coast of Africa. Proceedings Zoological Society of
er one n for nth pp. 411-413, 2 figs.
paps: "a Bradfor
893. rious Breadwinners of the Deep. Cosmopolitan Magazine,
Vol. 15, p. 756, fig.
Humboldt, PN par de.
1826. Essai Politique sur 1’Isle de Cuba, pp. 361-364.
1833. Recueil d’Observations de Zoologie et Anatomie Comparée, Vol.
II, pp. 192-193 in Voyage de Humboldt et Bonpland,
Deuxième Partie. Paris
1860. Personal Narrative of hens els to the Equinoctial Regions of
Ameri uring the Years 1779-1804. Translated and edited
by Thomasina Ross. Vol. IIT, pp. 182-186
Jardine, John.

1866. Description of the Neighborhood of Somerset, Cape York,
Australia. Journal Royal Geographical Society, London, Vol.
36, pp. 7
Kent, W. Saville

1893. The Greit Barrier pats of STRT its Products and Poten-

tialities, p. 321. Lon
Lacépède, B. G. E.

1829. Histoire Naturelle des Poissons, tome III, p. 490. Paris.
MacGillivray, John

1852. Narra tive of the i of H. M. 8. ee VoL O,
pp. 21-22. ndon
(Peter Martyr).

1504. gs rang de Tutta la Navigatione de Re de Spagna et de le
et Terrene Reena Trovati, chapter XV. Venezia,
a by Angelo Trivigian

524 THE AMERICAN NATURALIST -= Mo LM

_ 1507. Paesi Nouamente Retrouati et Nous Mondo da Alberico Vesputio
Florentino Intitulato, chapters “` LXXXIITI-CXIIII,
Peter Martyr.
1511. De Orbe Novo, Decade I, Liber 3; Decade VIII, Liber 8.
Seville. (Translated by F. A. MacNutt, and published by
Putnam, 2 Vols., New York, 1912.
PIF A. S:
913. A Naturalist in Cannibal Land, p. 205. London.
Ee Charles Theodor
1777. A New and Boei System of Geography; Country of Natal,
Vol. I, p. 388. London
Nieran, Johannes Eusebiu
Historia Naturæ, ‘hs XVI. Book XI, Chap. IV, pp. 250-
55, fig. Antwerpie
ssa! neues E.
Crusoe’s Island: A Bird Hunter’s Story, pp. 227-230, fig.
New York.
gad
167 Aerien Being the Latest and Most Accurate Description of
he New World, etc., pp. 49-50, fig. London
Oliver, S. pikis
1909. a Life of Philibert Commerson, D.M., Naturaliste du Roi.
n Old-World Story of French Travel sid Science in the Days
k Linnaeus, pp. 170-171. London
m y Valdez, Gonzalo Fernandez de. :
1547. Coronica de las Indias, pp. evii and 107. Salamanca.
Poey, F'elipe.
e Memorias ge la sr Natural de la Isla de Cuba,
o II, pp. 249-252. Haban

Pollen, a e
1874. Les Péches a hg gaan et ses Dépendances. In Faune de
Mada, Bees IV, by F. P. L. Pollen and D. ©. van Dam, pp.

2-3. Leiden.
Ramusio, Gisvansd Batista.
1606. Della Historia dell’ Indie, Liber XIII, Cap. X, pp. 133-134.
Venetia
Rivers, W. H. R.
1904. Reports Cambridge oe Expedition to Torres
Straits: Vol. V, p. 149.

Robinson, Phi
1883 Fishes of Fancy—Their Place in Myth, Fable, Fairy Tale, and
olk-Lore. Handbook Great International Fisheries Exhibi-
tion, pp. 22-23. London
Roth, H. Lin

g.
1889. Use of Sucker-fishes in Fishing. Nature ,Vol. 39.
ora W.E
1. Wood: its Search, ayers and Preparation. North Queensland
Ethnography, Bull etin 3

No. 629] STUDIES IN ECHENEIS OR REMORA 525

Salt, Henry.
1814. A Voyage to Abyssinia and Travels into the Interior of that
Country Executed under the Orders of the British Govern-
ment in the Years 1809 and 1810, p. 49. London.

Sclater, P. L.
1889. Use of`the Remora in Fishing. Nature, Vol. 39, p. 295.
Semon, Richar
1899. er the Australian Bush and on the Coast of the Coral Sea, p.

1903. im ppo Busch und den Küsten des Korallenmeeres.
age, Leipzig.
Sibree, aa
1915. A Naturalist in ——— p. 311. London.
a Andre
A dae au Cap de Bonne Maperanca: pp. 431-432. Paris.
mie Fein P
1903 Cuciatsther Columbus: His Life, His Works, His Remains, Vol.
II, pp. 215-218, 318, 360-363, 439-498. New York.
Thomas, N. W.
1 ke of Australia, p. 90. London.
Townsend, C. H.
1915. ne Power of the Shark-Sucker’s Disk. Bulletin New York
Zoo Mn aimed Vol. XVIII, pP. 1281-1283, 3 figs.
Wills, James, transl
2 X Fishing Fish (From an Old en ET Manuscript). Antan-
anariyo Annual, Vol. VI, p. 123.
Winsor, Just
1892 Christopher Columbus and How He Received and Imparted the
t of Discovery, 5th ed., pp. 9-10, 18, 34-56, 293. Boston.

VARIABILITY IN FLOWER-NUMBER IN
VERNONIA MISSURICA RAF.

DR. H. A. GLEASON

New York BOTANICAL GARDEN

In studying the species of Vernonia in the western
states, the writer was impressed as early as 1903 by the
constancy with which the number of flowers in each head
of certain species agrees with the numbers of the Fibo-
nacci series. This appeared at the time to be particularly
true of Vernonia fasciculata, which normally presents 18
to 21 flowers. Extending his studies later to the species
of tropical America, he found a still closer agreement `
with the Fibonacci series in the species with fewer-flow-
ered heads, where the numbers 8 and 13 are repeated
with little or no variation in several species. Inspired
by the more recent studies of Stout and Boas,! who re-
ported a steady seasonal decrease in flower-number in
the heads of Cichorium Intybus, he again examined in
1918 a series of specimens’ of Vernonia missurica, the
only species native in the vicinity of Ann Arbor, Mich.,
and made careful determinations of the number of flowers
in every head of several plants, chosen from different
localities and habitats. The results of these studies are
presented here. ;

In the western part of its range, from Kansas to Illi-
nois, Vernonia missurica is essentially a prairie species
and is seldom or never found in woods or swamps. In
Indiana it is excessively common on rolling hills in clay
sou, preferring land formerly wooded but now used for
pasture, where it is apparently avoided by live-stock.
In the extreme eastern end of its range, in southeastern

1 Stout, A. B., and Boas, H. M., ‘‘Statistical Studies of Flower Number
per Head in Cichoriwm Intybus; Kinds of Variability, Heredity and Effects
of Selection,’’ Mem. Torrey Club, 17, 334-458, pl. 10-13 + f. 1, 1918.

526

No. 629] FLOWER NUMBER IN VERNONIA . 527

Michigan and adjacent Ontario, it is typically a species
of moist, cleared, uncultivated bottom-lands along the
rivers and lakes and occasionally is found in low woods
as well, where it is taller and more slender, with rela-
tively lax, irregular, few-headed inflorescences. Judged
from herbarium evidence from all parts of its range, the
mode of its flower-number seems to be 34. Since flower-
numbers deviate more freely from the Fibonacci series
in the higher numbers and since the species has such a
wide range and such a wide variation in its habitat, it
may naturally. be expected that it will present a wider
variation in flower-number than many other species of
restricted range and habitat and with smaller heads.
The inflorescence of the species may be described in
some detail. It is strictly cymose. The main axis of the
plant is unbranched and rises to a height of one to two
meters. The lower nodes bear full-sized foliage leaves
and are separated by internodes of fairly uniform length,
while the upper internodes are abbreviated and bear re-
duced or bracteal leaves only. The main axis terminates
in a single head on a short peduncle, and this head is the
first, or among the first, to bloom. From the axils of the
uppermost leaves branches appear which are in turn
terminated by a single head on a longer peduncle, which
therefore overtops the primary head. The next nodes,
subtended by larger leaves, bear short simple leafless
cymes of 2-5 heads. Below these, leafy branches appear
from the lower nodes, and these bear single primary ter-
minal heads, axillary heads on longer peduncles, and
simple cymes in the same order in which they appear at
the summit of the main axis. The uppermost of these
leafy branches, since they arise relatively near the sum-
mit of the plant, also overtop the main axis and with it
produce a more or less flattened or depressed corymbi-
form cluster. ‘The lower lateral leafy branches are suc-
cessively shorter, bear fewer heads which bloom later,
and tend to produce in conjunction with the upper ones
a more or less cylindrical cluster. Finally, the lowest

528 THE AMERICAN NATURALIST (Von. LIII

nodes bear short and frequently undeveloped lateral
branches, which usually appear so late in the season that
none of their heads, or only a part of them, open their
flowers before frost. In many plants only the uppermost
nodes bear branches at all, and in such cases the inflores-
cence is flattened or depressed. Every plant normally
bears the primary terminal head and few or several sub-
terminal heads and simple cymes from the uppermost
axils. The number of heads is at a minimum in shady
situations. The middle nodes bear floriferous branches
only on large plants of favorable situations, where there
is sufficient light and the plants are not crowded for space.

A single cyme consists of two or more heads on pedun-
cles 1-3 em. long with subulate bracts. Each peduncle
is usually accurately curved and leaves the straight axis
at a prominent angle, so that the true terminal heads are
easily recognized. The usual number of heads in each
cyme is two to five and the maximum number observed is
nine. A cyme of two heads consists of a terminal head
and an inferior lateral head. A cyme of three heads con-
sists of the terminal and (a) two inferior lateral heads or
(b) an inferior two-headed cluster. ‘A cyme of four
heads presents the usual terminal head and (a) three
inferior lateral ones on separate peduncles or (b) one
lateral head and one two-headed cluster. One of five
heads has the true terminal and (a) a single inferior
lateral head and a three-headed cyme of either of the
types mentioned above or (b) two two-headed clusters.
Cymes of greater numbers of heads have the same gen-
eral structure, of a single terminal head with various
combinations of single inferior heads, two-headed clus-
ters and three-headed clusters.

Three types of variation were looked for in examining
the species: (1) a variation between the heads of each
eyme, possibly correlated with their position, whether
terminal or inferior; (2) a variation between different
floriferous branches of the same plant, possibly corre- -
lated with the amount of available nourishment; and (3)

No. 629] FLOWER NUMBER IN VERNONIA 529

a general variation between different individuals, pos-
sibly correlated with the size and vigor of the plant and
therefore indirectly with the habitat.

1. Within a single cyme of 2-6 heads, the terminal
head is usually the largest. In larger cymes of 7—9 heads
some of the secondary terminal heads, ending the lower
lateral branches of the leafless cluster, are frequently
larger than the primary terminal head.

Table I exhibits the number of flowers in the terminal

TABLE I
RELATION OF FLOWER-NUMBER TO POSITION ON THE BRANCH
oe eee ee eee
| ee Number Largest | Average
1 | 46 ee | 46 | 46
> 59 5 53 49.6
> 56 5 53 51.0
. 57 8 55 51.4
3 54 4 50 | 48.5
p 53 5 51 | 49.2
19 | 57 12 54 51.1
n 55 13 55 49.5
. | 57 10 55 49.3
13 57 54 50
n 58 9 56 52.6
ve 18 58 53.1
ve 52 14 57 52.9
= 55 14 59 52.1
as 50 21 58 | 52.9
= 54 22 60 | 53.4
“ty | 54 16 56 L id
pis 54 1 56 ber me
= 52 15 60 | 52.9
oe 59 8 62 53.0
pi 57 5 55 53.8
ro 53 2 50 48.5
id 50 3 52 | 49.7
= 53 2 53 50.0
Pete eRe A We. 165 Oe

head and the average number in the other heads on each
of 25 floriferous branches from the same plant, the num-
bers beginning at the base. On 4 branches, numbers 2, 5,
7 and 23, the terminal head was defective or worm-eaten,
and these have been omitted in the table. On 22 branches
of the 25, the primary terminal head is larger than the
average of the other heads and the difference may be as

530 THE AMERICAN NATURALIST [Vou. LII

much as 9.4. On 15 branches the primary terminal is
actually the largest head on the branch. In 3 cases the
primary terminal is smaller than the average, and in 10
cases it is exceeded in size by one or more of the lateral
heads. It will be noticed that these conditions occur only
on branches with numerous heads, where the terminal
heads of certain individual cymes tend to raise the aver-
age. In fact, on those branches which bear a total of
less than ten heads, and in which there are accordingly
fewer chances for large secondary terminal heads, the
average sizes of the two classes are 55 and 50.8 and with
two exceptions (branches nos. 24 and 27) the primary
terminal is actually the largest head on the branch. On
branches with a total of 10-20 heads the averages are 55
and 51.9 and the primary terminal is actually the largest
in only two fifths of the branches. In the two cases with
over 20 heads the averages are 52 and 53.2 and the pri-
mary terminals are conspicuously exceeded in size by
some of the other heads. Since the heads of each cyme
differ but little in age, the variation in their size may
possibly be due to difference in the amount of food-stuff
ór water available, by which the terminal heads at the
end of a continuous axis are favored.

2. It has already been stated that the solitary heads
and the floriferous branches appear in basipetal order
and that those from the lowest nodes may not be suf-
ficiently developed to bloom before frost stops all further
growth. Table II shows the variation in flower-number
correlated with the position of the branch.

The table indicates a steady increase in the number of
abortive heads from the older branches at the summit
to the younger ones at the base. The greatest number of
heads are found near the middle of the series on the
longest lateral branches, which rise from the middle
internodes to a height equal to or surpassing the summit
of the stem. But the average number of flowers is re-
markably constant throughout, varying only from 50.7
to 52.9 for each set and, in general, reaching the maxi-

No. 629] FLOWER NUMBER IN VERNONIA 531

mum among the larger branches. It is obvious that
there is very little relation between position and flower-
number and the same conclusion is supported by the data
from other plants.

TABLE II
RELATION OF FLOWER-NUMBER TO POSITION OF BRANCH
Number of Heads Number of Flowers
Branch 5 A Average by Groups
oN cae Fertile | Total | High | Low /Average| '
|
1 0 10 2 12 | 46 | 46 | 46.0
2 0 11 0 11 | | Fertile heads 3.2
3 1 16 6 23 59 | 47 51.2 Abortive heads 10
4 1 10 6 17 56 47 51.8 Flower-number 50.7
5 0 3 0 3 | |
6 a we oT ey er | aso |
7 1 8 6 15 | 53 | 47 | 49.5 | Fertile heads 8
8 0 5 5 10 | 54 | 47 | 49.6 | Abortive heads ?. 8
9 Ste 7 6 | 16] 53 | 48 | 498 | Flower-number 50.8
10 0 9 13 22 57 48 61.5 |
11 0 6| 14| 20] 55 | 45 | 499]
12 0 9 11 20 57 45 50.0 | Fertile heads 13.4
13 3 5 9 17 57 47 51.1 | Abortive heads 6.2
14 1 4 10 15 58 48 53.1 | Flower-number 51.6
15 0 oe 19 26 58 47 53.3
16 a pi 1 15 16 57 49 52. ‘
17 0 4 15 19 59 47 52.3 | Fertile heads 18.8
18 ł 5 22 28 58 46 52.8 | Abortive heads 3.8
ee 7 23 30 60 47 53.4 | Flower-number 52.8
20 | li 2 ibd 20 57 47 52.5
21 | 2 2 15 19 56 48 51.
22 3 1 16 20 60 49 52.8 | Fertile heads 12.4
23 3 0 8 11 58 47 53.0 Abortive heads 1.2
24 0 1 9 10 62 48 53.7 | Flower-number 52.9
25 0 2 6 8 57 52 54.3
|
26 | 1 0 3 4 53 47
27 0 0 + 4 52 46 49.8
28 0 p 3 of BR | SE OE aile bosk 2.1
29 0 0 2 2 52 50 51.0 ~
Abortive heads 0
30 0 0 1 1 Fl ber 51.0
31 0 0 1 1 | 52 ower-number .
32 0 0 1 1 | 49
B-r 0 0 1 1 | 56
Total...| 21 | 145 | 278 | di4 | 62 | 45 | 521

3. The number of maturing heads and the minimum,
maximum, and aver rage number of flowers per head in 22
plants is exhibited in Table III. Of these plants, num-
bers 1-18 inclusive were collected from a variety of

532 THE AMERICAN NATURALIST [Von. LII

habitats and stations, in shade and sun, and in relatively
wet and relatively dry soils. They show in every case a
small variation within each plant, but a great variation
between different plants, the averages ranging from 29.3
to 52.1. It happens that the plant with the largest num-
ber of heads also presents the highest flower-number,
but in general there is no correlation between: them, and

TABLE III
VARIATION IN FLOWER-NUMBER ON DIFFERENT PLANTS
Number of Flowers
Plant Number Heads Soe Ga ie eo en ay
High Low | Average
1 278 62 45 52.1
2 26 43 37 40.0
3 26 41 34 31.3
4 15 38 29 31.9
5 14 36 27 31.9
6 31 40 31 36.4
7 19 45 37 42.2
8 25 39 $2 35.4
9 95 37 | 22
10 62 35 | 26 29.7
11 43 39 28 34.2
12 15 51 | 45 47.7
13 rd 36 | 32 34.0
14 11 45 | 36 40.5
15 3 46 | 44 45.0
16 19 38 34 35.5
17 3 39 35 37.0
18 33 34 26 29
19 126 38 26 32.5
20 98 37 26 32.2
21 61 38 27 32.0
22 74 36 25 32.7

the third highest average is presented by the plant with
the smallest number of heads.

The last four plants, numbers 19-22, were collected
from the same station and grew under similar environ-
mental conditions in the usual (and for the region prob-
ably also the optimum) habitat of the species. They also
-present very similar averages in their flower-number.
Others of the same station were also examined and
demonstrated that essentially the same averages were
repeated throughout the group. |

From an examination of this table and from additional

No. 629] FLOWER NUMBER IN VERNONIA 533

experience with the plants, the writer is led to the ten-
tative idea that two sets of factors, which may be en-
vironmental, or hereditary, or both, act on the plants
independently, one determining the number of heads pro-
duced and the other the average number of flowers in
each, so that there may result plants with many large
heads (as no. 1), many small heads (no. 10), few large
heads (no. 15), or few small heads (nos. 5, 18).

Table 4 shows the distribution of flower-numbers for
all the heads of five plants. In each case the curves show
a close relation to the main or secondary numbers of the
Fibonacci series, 55, 29, 29, 34 and 34 respectively, al-
_ though in only two cases do the modes fall precisely on
these figures. Plant 1 shows a rather close grouping of
heads just short of 55, plant 9 has over half of the heads
grouped at 28-30, and plant 11 has almost half grouped
at 33-35. It is scarcely to be expected that the series will
be followed closely with such large numbers of flowers;
in fact, Stout has demonstrated that there is no relation
whatever to the Fibonacci series in the heads of Cichor-
ium Intybus. Since the numbers were determined in
every case by counting the mature achenes, the numbers
should fall somewhat below the Fibonacci series, rather
than above them, because of the possibility of some
flowers not setting seed.

The plants used for these five tabulations were selected
merely because of their large number of heads, which
offer better data for developing a representative curve.
Plants 20-22, with large numbers of heads also, agree
closely with plant 19. A moment’s inspection of the
averages for the other plants, as shown in Table III,
shows that in many cases, such as plants 2 and 3, they
could not agree closely with the Fibonacci series, or that
an apparent agreement might be fictitious if based on
plants with a few heads only, as numbers 12 and 13.

534 THE AMERICAN NATURALIST [Vor. LIII

TABLE IV
DISTRIBUTION OF FLOWER-NUMBERS
Plant 1
No. of heads... 3- 7 15 16 26 23 30 31 38 24 241315 7 3.2 0.1
No. of flowers. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62

Plant 9
Nos Of bonds: L 0-1 9-510 16 26 FSP 6 4 Bie dod: 1
No. of flowers. 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Plant 10
NO of Reade? 2° P4616 0 90 0 t21
No. of flowers. 26 27 28 29 30 31 32 33 34 35

Plant 11
No. Of kdn FEO FTT 8 T eo +g 2
No. of flowers. 28 29 30 31 32 33 34 35 36 37 38 39

Plant 19
Nó. of hadi.. G3 0 4°15 (1S 12 olosi o sg 1
No. of flowers. 26 27 28 29 30 31-32 33 34 35 36 37 38

SUMMARY

1. The number of flowers in each head is greatest for
the terminal heads of each cyme.

2. Otherwise the number of flowers in each head is ein
tively constant for each individual plant.

3. There is a great variation between individuals but,
in those plants with numerous heads, the mode falls on
or near one of the main or secondary numbers of the
Fibonacci series.

DARWIN’S CONTRIBUTION TO THE KNOWLEDGE
OF HYBRIDIZATION

HERBERT F. ROBERTS

UNIVERSITY OF MANITOBA

THe period from 1859 until the re-discovery of Men-
del’s papers in 1900 was so strongly colored by the views |
of Charles Darwin, and so dominated by the magnitude of
his work, that it sometimes seems as though originality
and initiative had been abandoned, and as though, so far
as evolution were concerned, the scientific world had re-
mained content simply to quote Darwin.

It is the purpose of the present paper to present the
contributions of Darwin to the knowledge of hybrids. To
this end it seems desirable, so far as possible, to let Dar-
win’s words speak for themselves, and hence, although
the paper may seem burdened with extracts, yet, for those
interested in tracing the history of ideas in genetics, it
will perhaps be of service to assemble such a résumé of
Darwin’s work and thought in the field of hybridization.
Brought together in such a way, an author’s contribution
may be more successfully valuated at leisure by those who
may be interested. The writer has, therefore, sought to
bring together, in somewhat connected and coherent form,
the various views, conclusions and experimental data on
the ee gs of hybrids and hybridization, found in Dar-
win’s different writings.

It is also the purpose of the writer to bein into relief
certain fields of investigation opened by him, but in which,
so far as appears, there has been little or no research
since his time.

On November 24, 1859, appeared the first edition of
‘‘The Origin of Species (1), antedating by seven years,
the appearance of the papers of Mendel.

; 535

536 THE AMERICAN NATURALIST [Von. LIII

One of the primary questions concerning crossing that
interested Darwin was the matter of sterility and fertility
in hybrids. Investigators before Darwin’s time had been
to a considerable extent obsessed by the species question,
which crossing was supposed to solve. If a cross suc-
ceeded, or produced fertile offspring, it argued that the
parent forms were ‘‘varieties.’’ If the cross failed, or if
its offspring were sterile, it demonstrated that they were
‘*species.’’ With the sole exception of Sageret (2), none
of the earlier hybridists seems to have formed anything
like the unit-character conception, and with the sole ex-
ception of Naudin and Darwin, no scientific theory was
conceived of which might explain the modus operandi of
amphimixis in the case of hybrids.

By Darwin, the question of hybridization, while indeed
for the most part, taken up more or less conventionally,
received, nevertheless, broader treatment. To begin with,
Darwin held that the inability of species to cross ‘‘is
often completely independent of their systematic affinity,
that is, of any difference in their structure or constitution,
excepting their reproductive systems’’ (la, 2:14).

So that, even as early as the writing of the ‘‘Origin of
Species,’’ Darwin is seen to maintain that the susceptibil-
ity of plants to crossing stood in no relation to the degree
of their resemblance to either parent, and that ‘‘the facil-
ity of making a first cross between any two species is not
always governed by their systematic affinity or degree of
resemblance to each other’’ (1a, 2:16).

This fact, he adds, is demonstrated by the case of re-
ciprocal crosses, alluding here to the relative facility of
making the cross, according as the one or the other species
is used as the male or the female.

Occasionally he says there is ‘‘the widest possible dif-
ference in the facility of effecting a union. The hybrids,
moreover, produced from reciprocal crosses, often differ
in fertility? (ibid.).

Darwin again later in ‘‘Animals and Plants under
Domestication,’’ refers to the matter as follows:

No. 629] DARWIN AND HYBRIDIZATION 587

Why should some species cross with facility, and yet produce very
` sterile hybrids, and other species cross with extreme difficulty, and yet
produce fairly fertile hybrids? Why should there often be so great
a difference in the result of a reciprocal cross between the same two
species? (la, 2:17).

Darwin comments frequently in the ‘‘Origin of Spe-
cies,” upon the fact that the hybrids produced from re-
ciprocal crosses often differ in fertility, and that while
two species may be difficult to cross, there is no strict
parallelism between the difficulty of effecting the cross
and the degree of sterility of the hybrids resulting there-
from.

As Darwin observes, differences in the results in respect
to the relative ease of making reciprocal crosses had been
previously noted by Koelreuter, who found, after two
hundred trials, continued for eight years, that while Mira-
bilis jalapa could easily be fertilized by M. longiflora,
the reverse cross could not be effected.

With regard to the difference in the facility with which
reciprocal crosses can be made, there may be some fun-
damental resemblance between this fact and the ease with
which reciprocal grafts can be made, wherein Darwin in-
stances the fact that the currant can, although with dif-
ficulty, be grafted upon the gooseberry, while the recip-
rocal graft can not be made. Certainly the well-estab-
lished facts of somatic segregation followed by germinal
‘‘mutation’’—so-ealled, should sufficiently indicate that
the behavior of the somatic and of the reproductive cells
should not be regarded as being so sharply separated as
is usually done in genetic studies. At all events, the prob-
lem as to the reason for the relative differences in the re-
spective facility of making reciprocal crosses, as well as
the further one of the differences as in the case of mule
and hinny, between the respective products of reciprocal -
crosses, are questions that have been but very little inves-
tigated since Darwin’s time, and demand thorough ex-
-ploration.

Since the advent of Mendelian studies in 1900, it has

538 >. THE AMERICAN NATURALIST [Von. LIII

been rather conventionally and very loosely assumed that
reciprocal crosses are invariably identical in type. That
such is not necessarily the case, Darwin’s early observa-
tions should suffice to indicate.

The problem of the fertility of selfed and crossed
plants engaged Darwin’s close interest in forty-one cases
belonging to twenty-three species. The ratio of the fer-
tility of the crossed to that of the self-fertilized plants was
found to be as 100:60. In another experiment to deter-
mine the relative fertility of flowers when crossed or
selfed, the ratio in thirty cases belonging to twenty-seven
species was as 100:55.

There is no evidence, Darwin finds, That the fertility of plants goes
. on diminishing in successive self-fertilized generations, and no close
correspondence, either in the parent plants or in the successive genera-
tions, between the relative number of seeds produced by the crossed
and self-fertilized flowers, and the relative powers of growth of the
seedlings raised from such seeds (1b, 327).

Darwin’s investigations were directed quite extensively
to the question of self-fertility in plants, a field which
bears strongly upon our knowledge of heredity, but in
which likewise comparatively little experimental work
has been done since his time. As the result of his own
studies, supplemented by those of Hildebrand and Fritz

üller, he was able to say:

We may therefore confidently assert, that a self-sterile plant can be
fertilized by the pollen out of any one out of a thousand or ten thou-
sand individuals of the same species, but not by its own (ibid., 347).

Regarding the cause of sterility or inability to accept
fertilization, we are as greatly at a loss for an explanation
to-day as was Darwin. As Darwin well states it:

The veil of secrecy is as yet far from lifted; nor will it be, until we
can say why it is beneficial that the sexual elements should be differ-
entiated to a certain extent, and why, if the differentiation be carried
still further, inquiry follows. It is an extraordinary fact that with
many species, flowers fertilized with their pollen are either absolutely

No. 629] DARWIN AND HYBRIDIZATION 539

or in some degree sterile; if fertilized with pollen from another ftower
on the same plant, they are sometimes, though rarely, a little more
fertile; if fertilized with pollen from another individual or variety of
the same species they are fully fertile; but if with pollen from a distant
species they are sterile in all possible degrees, until utter sterility is
reached. We thus have a long series with absolute sterility at the
two ends; at one end due to the sexual elements not having been dif-
ferentiated, and’ at the other end to their having been differentiated
in too great a degree, or in some peculiar manner (ibid., 455).

The questions which Darwin raises in this connection
are as follows (p. 458) :

1. Why the individuals of some species profit greatly,
others very little by being crossed.

2. Why the advantages from crossing now seem to ac-
crue exclusively to the vegetative and now to the repro-
ductive system, although generally to both.

. Why some members of a species should be sterile,
while others are entirely fertile with their own pollen.

4. Why a change of environment or of climate should
affect the sterility of self-sterile species.

5. Why the members of some species should be more
fertile with the pollen from another species than with
that of their own.

Regarding the general matter of sterility in hybrids,
Darwin comments as follows:

It is notorious that when distinct species of plants are crossed they

r. This unproductiveness varies in different species up to sterility
so complete that not even an empty capsule is formed (1b, 463).

It is also notorious that not only the parent species, but the hybrids
raised from them are more or less sterile, and that their pollen is in a
more or less aborted condition. The degree of sterility of various
hybrids does not always strictly correspond with the degree of difficulty
in uniting the parent forms. When hybrids are capable of breeding
inter se, their descendants are more or less sterile, and they often
become still more sterile in the later generations (ibid.

With the majority of species, flowers fertilized with heit own pollen
yield fewer, sometimes much fewer seeds, than those fertilized from
another individual or variety (ibid., 464).

540 THE AMERICAN NATURALIST [Vou. LIN

As the result of his investigations regarding sterility
to pollen, Darwin was able to render at least one service,
that of removing the obsession which had so long af-
flicted the study of the hybrid question, viz., the variety-
species discussion. He says:

It can thus be shown that neither sterility nor fertility affords any
certain distinction between species and varieties. The evidence from
this source graduates away, and is doubtful in the same degree as is the
evidence derived from other constitutional and .structural differences
(la, 2: 4)

The question of the chemical and cytological basis for
sterility or non-receptivity to pollen, remains still in part
an open field for the investigator.

One of the most important questions from our present-
day point of view which Darwin investigated was the
relative vigor of first-generation hybrids as compared
with that of their parents. The following allusions occur
in the ‘‘Origin of Species.’’

Darwin comments on the fact that crosses between in-
dividuals of the same species, where they differ to a cer-
tain extent, give increased vigor and fertility, while close
fertilization, long continued, almost always leads to phys-
ical degeneracy, and remarks:

We know also that a cross between the distinct individuals of the
same variety, and between distinct varieties, increases the number of the
offspring, and certainly gives to them increased size and vigor (la,
2: 296).

Darwin havoughty investigated, as is well known, the
comparative relation of the offspring of crossed to those
of selfed plants with respect to vigor.

I have made so many experiments, and collected so many facts, show-
ing on the one hand that an occasional cross with a distinet individual
or variety increases the vigor and fertility of. the offspring, and on the
other hand that very close interbreeding lessens their vigor and fertility,
that I can not doubt the correctness of this conclusion (2a, 2

No. 629] DARWIN AND HYBRIDIZATION 541

Again, from both plants and animals, there is the clearest evidence
that a cross between individuals of the same species, which differs to a
certain extent, gives vigor and fertility to the offspring, and that close
interbreeding, continued during several generations between the nearest
relations, if these be kept under the same conditions of life, almost
always leads to decreased size, weakness or sterility (la, 2: 27).

In ‘‘Cross and Self Fertilization,’? Darwin again dis-
cusses the effects of crossing as follows, expressing the
view:

Firstly, that the advantages of cross-fertilization do not follow from
some mysterious virtue in the mere union of two distinct individuals,
but from such individuals having been subjected during previous gen-
erations to different conditions, or to their having varied in a manner
commonly called spontaneous, so that in either case their sexual ele-
ments have been in some degree differentiated; and secondly from the
want of such differentiation in the sexual elements (1b, 443).

A eross with a fresh stock or with another variety seems to be always
beneficial, whether or not the mother plants have been intercrossed or
self-fertilized for several previous generations (1b, 444).

Darwin also remarks upon the greater power of the
eross-fertilized plants in his experiment to stand expo-
sure, the crossed plants enduring sudden removal from
greenhouse to out-of-door conditions better than did the
self-fertilized, and also resisting cold and intemperate
weather conditions more successfully. This was the case
with morning glory and with Mimulus.

The offspring of plants of the eight self-fertilized generations of
Mimulus crossed by a fresh stock, survived a frost which killed every
single self-fertilized and intercrossed plant of the old stock.

Independently of any external cause which could be detected; the
self-fertilized plants were more liable to premature death than the
crossed (ibid., 290).

Out of several hundred plants in all involved in the ex-
periment, only seven of the crossed plants died, while at
least twenty-nine of the self-fertilized were thus lost.

With regard to time of flowering, in four out of fifty-

542 THE AMERICAN NATURALIST [Von. LIII

eight cases, a crossed plant, in nine cases a selfed plant,
flowered first.

Darwin broached the view that the increased vigor of
first-generation hybrids was chiefly due to the forms used
in the cross having been exposed to somewhat different
conditions of life. He also contended that his experi-
ments proved that

If all the individuals of the same variety be subjected during several
generations to the same conditions, the good derived from crossing is
often much diminished or wholly disappears (la, 2: 270

This statement appears to be an obiter dictum of Dar-
win’s to the support of which he does not adduce direct
experimental evidence.

Again he says:

Anyhow my experiments indicate that crossing plants which have
been long subjected to almost though not quite the same conditions, is
` the most powerful of all the means for retaining some degree of dif-
ferentiation in the sexual elements, as shown by the superiority in the
later generations, of the intercrossed over the self-fertilized seedlings
(1b, 450).

We know, he says, that 1 a plant prepeeated for some generations in
ancther garden in the same district serves as a first stock, and has
high fertilization powers (ibid.).

The importance of this view has yet, so far as the
writer knows, to be thoroughly re-investigated under com-
pletely controlled conditions.

It was Darwin’s view, as the result of his experiments,
that the increased vigor of intercrossed plants is due to
the constitution or nature of the sexual elements, which
condition he took to be of the general nature of differen-
tiation due to the action of environment.

It is certain, he says, that the differences are not of an external
nature, for two plants which resemble each other as closely as the
individuals of the same species ever do, profit in the plainest manner
when intererossed, if their progenitors have been exposed during several
generations to different conditions (1b, 270).

No. 629] DARWIN AND HYBRIDIZATION 543

Darwin asserts that there is not a single case in his ex-
periments,

Which affords decisive evidence against the rule, that a cross between .
plants, the progenitors of which have been subjected to somewhat
diversified conditions, is beneficial to the offspring (ibid.,

The fact that increased vegetative vigor in first gen-
eration hybrids was also sometimes accompanied by di-
minished fertility was likewise observed by Darwin,

For it deserves especial attention that mongrel animals and plants,
which are so far from being sterile, that their fertility is often actually '
augmented, have, as previously shown, their size, hardiness and con-
stitutional vigor generally increased. It is not a little remarkable that
an accession of vigor and size should thus arise under the opposite con-
tingencies of increased and diminished fertility (1e, 2: 108).

In the case of Darwin’s experiments to determine the
relative effects upon vigor of selfing and crossing, respec-
tively, the data were determined chiefly with respect to
height and weight of the plants, which were grown on
opposite sides of the same pot in all instances.

Regarding the relative heights and weights of 292
plants, derived from a cross with a fresh stock, and of
305 plants either selfed or intercrossed, between plants
of the same stock and belonging to thirteen species and
twelve genera, Darwin says:

Considering all the eases... there can be no doubt that plants
rofit immensely, though in different ways, by a cross with a f
stock, or with a distinct sub-variety. He emphasizes further, It can
not be maintained that the benefit thus derived is due merely to the
plants of the fresh stock being perfectly healthy, whilst those gra
had been long intercrossed or self-fertilized had become unhealthy ;
in most cases there was no appearance of such unhealthiness (1b, 260).

Experiments were also made with plants belonging to
five genera in four different families. One of the most
interesting cases was that of a plant of marjoram (Ori-
ganum vulgare). The height of the crossed was to that of
the selfed plants as 100: 86.

544 THE AMERICAN NATURALIST ‘[ Vou. LII

They differed also to a wonderful degree in constitutional vigor. The
crossed plants flowered first, and produced twice as many flower-stems;
and they afterward increased by stolons to such an extent as almost
to overwhelm the self-fertilized plants (1b, 302).

Darwin holds that the inferiority of the selfed seed-
lings in height can have been in no way due to any mor-
bidity or disease in the mother plants; certainly, he main-
tains, no such theory of a diseased condition would in any
wise hold, in the case o

intererossing the individuals of the same variety or of distinct varie-
ties, if these have been subjected during some generations to different
conditions (1b, 445)

In four out of the five cases experimented with, the in-
tercrossing of flowers upon the same plant did not differ
in effect from" the strictest self-fertilization. Conclud-
ing, he says:

On the whole the results here arrived at... agree well with our
general conclusion, that the advantage of a cross depends on the
progenitors of the crossed plants possessing somewhat different con-
stitutions, either from having been exposed to different conditions, or
to their having varied from unknown causes in a manner which we in
our ignorance are forced to speak of as spontaneous (1b, 302).

Darwin’s experiments indicated as in the case of
heartsease and sweet pea, that

the advantage derived from a cross between two plants was not con-
fined to the offspring of the first generation (1b, 305).

Laxton’s varieties of sweet peas produced by crossing,
as Darwin says:

have retained their astonishing vigor and luxuriance for a considerable
number of generations (ibid.).

Darwin concludes:

As the advantage from a cross depends on the plants which are
erossed differing somewhat in constitution, it may be inferred as prob-

No. 629] DARWIN AND HYBRIDIZATION 545

able that under similar conditions, a cross between the nearest‘ relations
would not benefit the offspring so much as one between non-related
plants (ibid.).

Darwin finally also remarks in general:

Tt is interesting to observe ... the graduated series from plants
which, when fertilized by their own pollen, yield the full number of
seeds, but with the seedlings a little dwarfed in stature, to plants which,
when self-fertilized, yield few seeds, to those which yield none, but have
their ovaria somewhat developed,—and, lastly, to those in which the
plant’s own pollen and stigma mutually act on one another like poison
(le, 2: 119).

The relative weight and germinative energy of seeds
from crossed and from self-fertilized plants, was inves-
tigated by Darwin in the case of sixteen species, with the
result that the weight of the seeds of the former to that of
the latter was found on the average to be as 100:96. In
ten out of the sixteen cases the self-fertilized seeds were
either equal or superior to the crossed in weight, but in
six out of these ten, the plants raised from these selfed
seeds were greatly superior in height and in other re-
spects to those from the crossed seeds. In the matter of
germination of selfed and crossed seeds, the results were
conflicting. Darwin, however, discovered that, in general,
seedlings of greater constitutional vigor are obtained
when crossed by other individuals of the same stock, than
when pollinated by their own pollen.

n the cases of plants of fifty-seven different species,
belonging in all to fifty-two genera and thirty different
families, Darwin carried out the most extensive exper-
iment yet recorded, conducted for the purpose of deter-
mining the differences in size, between the offspring of
cross-fertilized and of close-fertilized plants.

e total number of the crossed plants amounted to
1101, and of the selfed plants to 1076. As a result, Dar-
win found that the plants derived from crosses between
different strains of the same species, were taller on the
average, than plants derived from erosses within the

546 THE AMERICAN NATURALIST [Von LIII

same strain, and taller in the latter case than in the case
of the offspring of. self-fertilized plants. The average
ratio of 620 crossed to 607 selfed plants in height, derived
from Darwin’s tables, was as 100: 86

From the fact that flower buds are in a sense distinct
individual plant units, which sometimes vary and differ
widely from one another, and yet, when on the same plant,
owing to the fact that the plant has come from the same
fertilized cell, rarely are widely differentiated, Darwin
reasons that the effects of intercrossing can be explained.
He says:

The fact that a cross between two flowers on the same plant does no
good or very little good, is likewise a strong corroboration of our.
conclusion ; for the sexual elements in the flowers on the same plant can
rarely have been differentiated, though this is possible, as flower buds
are in one sense distinct individuals, sometimes varying and differing
from one another in structure or constitution (1b, 444).

Hence, he concludes:

Thus the proposition that the benefit from cross-fertilization depends
on the plants which are crossed having been subjected to somewhat
different conditions, or to their having varied from some unknown
cause as if they had been thus se is securely fortified from all
sides (1b, 444).

Darwin comments also on the reversed situation, where
changes in the external condition result in sterility, for
which he seeks to find a logical connection with the condi-
tion induced by crossing.

On the one hand, slight changes in the conditions of life are favor-
able to plants and animals, and the crossing of varieties adds to the
size, vigor, and fertility of their offspring, so on the other hand, cer-
tain other changes in the conditions of life cause sterility; and as this
likewise ensues from crossing much modified forms or species, we have
a parallel and a double series of facts, which apparently stand in
close relation to each other (le, 2: 126).

Darwin’s view as to the reason for the good effects of
crossing was based upon the long prevalent opinion that,

No. 629] DARWIN AND HYBRIDIZATION 547

since animals, and hence presumably plants, profit from
changes in their conditions, that probably such changes
operate to affect the germ cells, or that in some way the
germ cells receive an extra stimulation on that account,
_which redounds to the benefit of the offspring (1c, 2:155

Darwin appears to hold the ill effects of close fertiliza-
tion to be due to the fact that the sexual elements in the
different flowers on the same plant have not differen-
tiated, while in his conclusion he appears to consider the
benefits of cross-fertilization to be due to the individuals
involved in the cross having differentiated through hav-
ing been exposed to different conditions.

Darwin frequently emphasizes the same view regarding
the differentiating effects of a new environment.

But hardly any cases afford more striking evidence how powerfully

a change in the conditions of life acts on the sexual elements, than

those already given, of plants which are completely ‘self-sterile in one

country, and when brought to another, yield even in the first genera-

tion, a fair supply of self-fertilized seeds (1b, 477), and again, . . . we

know that a plant propagated for some generations in another garden
as res

ized by any other individual of the same species but are altogether
sterile with their own pollen, become intelligible, if the view here pro-
pounded is correct, namely, that the individuals of the same species,
growing in a state of Nature near, have not really been subjected during
several previous generations to quite the same conditions (1b, 450).

When two varieties which present well-marked differences are crossed,
their descendants in the later generations differ greatly from one

a
obliteration of some of these characters, and to the Ses pala of
former ones through reversion; and so it will be, as we may

sure, with any slight differences in the constitution of ar sexual
elements (1b, 449).

With regard to the ill effects derived from self-f ase mad
tion, Darwin says:

Whether the evil from self-fertilization goes on inereasing during
successive generations is not as yet known, but we may infer from my

548 THE AMERICAN NATURALIST (Von. LILII

experiments that the increase, if any, is far from rapid. After plants

have a strictly analogous result with our domestic animals. The good
effects of cross fertilization are transmitted by plants to the next gen-
eration, and judging from the varieties of the common pea, to many -
succeeding generations. But this may merely be that crossed plants of
the first generation are extremely vigorous, and transmit their vigor
like any other character to their successors (1b, 438).

In this paragraph Darwin calls attention to a fact that
attracted little attention for a generation,—viz., the im-
mediate improvement due to a cross. Darwin was thus if
not the first to call sharply to attention, the matter of the
relatively increased size and vigor of first generation
hybrids, at least the first to subject the question to exper-
imental analysis.

So far as plant hybrids are concerned, Darwin’s mind
was chiefly occupied, as we have seen, not so much with
the fundamental theory of hybrids, as with the question
of sterility in hybrids and its inheritance. The general
question of what is the essential nature of hybridity, and
how and in what manner the characters are distributed in
the hybrid offspring, seems not to have come to an issue
with him

However, among the matters of interest to modern stu-
dents of genetics are his recognition of the general fact
of the intermediacy of F, hybrids, and of the occasional
complete dominance of one or the other set of parental
characters, together with the phenomena which he terms
‘‘reversion.’’ Regarding the former matter he remarks:

There are certain hybrids which, instead of having, as is usual, an
intermediate character between their two parents, always closely re-
semble one of them (1, 2:15).

In regard to the behavior of characters in crosses, while
admitting that, in the majority of cases, the hybrid off-
spring are intermediate between their parents, he recog-
nized that certain characters are incapable of fusion.

No. 629] ` DARWIN AND HYBRIDIZATION 549

When two breeds are crossed, their characters usually become inti-
mately fused together, but some characters refuse to blend, and are
transmitted in an unmodified state, either from both parents or from
one (le, 2:67).

As cases in point, Darwin cites the crossing of gray and
white mice, the offspring being pure white or gray, but
not intermediate, and the crossing of white, black and
fawn-colored Angora rabbits, in which the colors are sep-
arately inherited, and not combined in the same animal.
The non-intermediate character of the inheritance in the
case of turnspit dogs and ancon sheep is referred to, as
is also the inheritance in the case of tail-less, horn-less
breeds. Similar results in the case of stocks, toad-flax
and sweet peas are cited (1b, p.

Darwin (le, 44-45), in discussing what he called ‘‘pre-
potency,” was dealing in very many cases with that
which we now recognize as simple dominance. For ex-
ample, in the crossing of snap-dragons, Darwin found
that when the normal or irregular-flowered type was
crossed with the abnormal or regular-flowered type, the
former prevails in the first generation to the exclusion of
the latter. These 127 hybrid plants self-fertilized, yielded
in the second generation irregular to regular plants in the
ratio of 88 to 37. This is very close to the exact 3:1 ratio
which would be represented by the numbers 85:42. Dar-
win, however, simply regards it as a _

good instanee of the wide difference between the inheritance of a char-
acter and the power of transmitting it to the crossed offspring (1b, 45).

Darwin was thus quite unable, with the information
then available, to frame a satisfactory explanation for the -
various phenomena —_— under the name of ‘‘pre-
potency.’’

He makes one remark akve to prepotency, however,
that slightly grazes the present-day presence and absence
theory of Mendelian inheritance.

550 THE AMERICAN NATURALIST [Vou. LHT

We can seldom tell what makes one race or species prepotent over
another; but it sometimes depends on the same character being present
and visible in one parent, and latent or potentially present in the
other (1c, 2: 58)

The matter of sex-linked characters did not escape Dar-
win’s observation, alluding to cases where a son does not
inherit a character directly from his father, or transmit it
directly to his son, but receives it by transmission from a
mother who does not show it, and transmits it through his
non-affected daughter. Darwin observes:

We thus learn that transmission and development are distinct powers

(ibid.)

Respecting the matter of reversion, or what we should
now call recombination after segregation, Darwin’s utter-
ances are remarkable, especially in ‘‘ Animals and Plants
under Domestication.’’ In most cases he regards ‘‘re-
version’’ as the coming to light of a ‘‘latent’’ character,
as, €. g.,

hornless breeds of cattle possess a latent capacity to reproduce horns,
yet when crossed with horned breeds they do not invariably produce
offspring bearing horns (le, 2:44).

Darwin deserves credit for strictly contesting the point
of view then widely current, that the longer a character is
handed down by a breed, the more fully it will be con-
tinued in transmission. Discussing some of the cases, he
says (1c, 2:37): 2

In none of these nor in the following cases, does there appear to be any
relation between the force with which a character is transmitted and
the length of time during which it has been transmitted.

The basis for such a view, that the longer a breed is
handled and the more it is selected, the more homozygous
it becomes, was not scientifically known in Darwin’s time,
but Darwin actually perceived that the mere repeated act
of selection itself, whatever else might be involved, would

No. 629] : DARWIN AND HYBRIDIZATION 551

not increase the potency of transmission, or eliminate be-
yond question the liability to reversion.

Darwin considered it doubtful whether, as was then
popularly supposed, the length of time during which a
character had been inherited, had any influence on its
fixedness, and concluded from the fact that when wild
species which have remained so for ages, are brought into
cultivation, they immediately begin to vary, that no char-
acter by long inheritance can be considered as absolutely
fixed (1c, 2:56).

In this work, more than elsewhere, Darwin devoted
himself particularly to the question of the meaning of
inheritance in hybrids. The question always demanding
explanation was the reason for the reappearance after
the first generation of a hybrid of a parental, or even of
an ancestral form, a phenomenon then called ‘‘rever-
sion,” including, as Darwin says:

all cases in which an individual with some distinguishable character, a
race or species, has, at some former period been crossed, an el
acter derived from his cross, after having disappeared during one or
several generations, suddenly reappears (le, 2:2).

Darwin, at the outset, merely comments on the result
of crossing as follows:

In considering the final result of the commingling of two or more breeds,
we must not forget that the act of crossing in itself tends to bring back
long-lost characters not proper to the immediate parent form (le,
2: 64

It was noticed that from three to eight generations
were usually required before a breed derived from a
cross comes to be considered free from danger of rever-
sion. What constituted the machinery to bring about re-
version remained, but for Mendel’s as yet undiscovered
researches, absolutely unknown. The state of knowledge
in that regard is well exemplified by Darwin’s remark,

That the act of crossing in itself gives an impulse towards reversion, as
shown by the reappearance of long-lost characters, has never, I believe
been hitherto proved (1e, 2:13).

552 THE AMERICAN NATURALIST [Vou. LIII

Darwin recognized, as did most of the breeders before
Mendel, that

As a general rule, crossed offspring in the first generation are nearly
intermediate between their parents, but the grandchildren and succeed-
ing generations continually revert in a greater or lesser degree, to one
or both of their progenitors (1e, 2:22).

From eases of intermediacy, Darwin proceeds to dis-
cuss what we should call cases of dominance, and finally
eases in which the offspring in the first generation are
neither intermediate nor uni-parental in type, but in
which there is vegetative splitting, or somatic segre-
gation:

In which differently colored flowers borne on the same root resemble
both parents, . . . and those in which the same flower or fruit is striped
or blotched with the two parental colors, or bears a single stripe of the
color or other characteristic quality of one of the parent forms
(le, 2: 69).

It is interesting to see how Darwin now undertook, in
the absence of experimental evidence, to devise a scien-
tific solution for the reappearance of parental characters
in the second generation of the offspring. Taking Nau-
din’s idea of segregation or ‘‘disjunction’’ of the ele-
ments of the species, he concludes as follows:

If . . . pollen which included the elements of one species happened
to unite with ovules including the elements of the other species, the
intermediate or hybrid state would still be retained, and there would be
no reversion. But it would, as I suspect, be more correct to say that
the elements of both parent species exist in every hybrid in a double
state, namely, blended together and completely separated (le, 2:23).

The above paragraph comes more nearly being a state-
ment of the true nature of the hybrid or heterozygote con-
dition as Mendel’s analysis has revealed it, than any other
account hitherto published.

Combining this with the following statements, we have
very nearly the same idea which Mendel’s theory in-

No. 629] DARWIN AND HYBRIDIZATION 553

volves, based, however, upon Darwin’s theory of pan-
genesis, whereby each cell was supposed to throw off
‘‘vemmules’’ which carried the characters to the repro-
ductive cells. He says:

The tendency to reversion is often induced by a change of conditions,
and in the plainest manner by crossing. Crossed forms of the firs
generation are generally nearly intermediate in character between their
two parents, but in the next generation the offspring commonly revert
to one or both of their grandparents, and occasionally to more remote
ancestors (le, 2: 383).

Darwin then assumes that in the hybrid there exist
two kinds of ‘‘gemmules’’ or character-carriers; viz.,
pure gemmules from each of the two parent forms, and
combined or hybridized gemmules as well, and proceeds
in the following statement, to give about as clear an ac-
count as we have to-day, of the cause for the reappear-
ance of the parental or homozygote forms.

. .. when two hybrids pair, the combination of pure gemmules de-
rived from the one hybrid with the pure gemmules of the same parts
derived from the other would necessarily lead to complete reversion of —
character, and it is perhaps not too bold a supposition that unmodified
and undeteriorated gemmules of the same nature would be especially
apt to combine.

Pure gemmules in combination with hybridized gemmules would lead
to partial reversion, and lastly, hybridized gemmules derived from both
parent-hybrids, would simply reproduce the original hybrid form. All
these eases and degrees of reversion incessantly occur (le, 2: 383).

The latter statement is virtually a statement of the con-
dition of things in heterozygosis, in principle as we have
it to-day. If we assume the ‘‘hybridized’’ gemmules to
represent the ‘‘Dr.’’ combination, we have the necessary
substitution.

Darwin’s theory was a natural corrollary to his doctrine
of pangenesis. It seems strange that with Naudin’s idea
of disjunction in hand, and with the phenomenon of segre-
gation in peas, noticed by five observers, all of whose ex-
periments Darwin remarks upon, that = did not

554 = THE AMERICAN NATURALIST [Vou. LIII

himself perform Mendel’s experiment. However, it is a
manner of special interest that à priori, in the absence of
experimental data, he should have come as near the prin- `
ciple of the Mendelian explanation as the ane passages
indicate.
BIBLIOGRAPHY
1. Darwin, Charles
The Origin of Species by Means of Natural Selection, or the
A of Favored rome in the Struggle for Life. Lon
don ed., New York,
b. 1877. The pees ‘of ee are fae fertilization in the Vegetable

Kingdom. New
c. 1900. The Vailition a aan and Plants under Domestication.
2d ed., New York.
2. Sageret, Augustin.
1826. Considérations sur la production des hybrides, or variantes, et
étés en général, et sur celles des Cucurbitaceés en par-
ticulier. Annales des Sciences Naturelles, nm priest

SHORTER ARTICLES AND DISCUSSION

DOES EVOLUTION OCCUR EXCLUSIVELY BY LOSS OF
GENETIC FACTORS?

In an extremely interesting article, Professor Duerden’ has re-
cently discussed certain aspects of evolution in the light of obser-
vations on ostrich farming. He shows that as regards most char-
acters the germ plasm of the ostrich is remarkably stable and yet
that quantitative variation as regards wing and toe characters is
occurring and is being utilized, in particular for a gradual ameli-
oration of the valued plume characters. He believes that the
quantitative variation in question has a factorial genetic basis, a
view which I see no reason to question. He holds that repeated
selection may probably extend the existing range of variation
downward, but not upward. In this last conclusion I can not `
concur. It rests, I believe, on too close adherence to the ‘‘pres-
ence-absence hypothesis.’’ It ‘assumes that minus variation
occurs only by loss of factors and further that factors once lost
ean not be recovered. I do not think that either of these assump-
tions will bear critical examination. Morgan has recorded, in
Drosophila, the occurrence of a reversed mutation by which col-
ored eyes were recovered in a white-eyed race, and on this ground
has questioned the validity of the entire presence-absence hy-
pothesis. I have found that in the piebald patterns of rats and
rabbits steady progress may be made by repeated selection in
changing the racial average either in a plus or in a minus direc-
tion. Genetic changes affecting the extent of the. pigmented
areas are clearly of frequent occurrence in such cases, precisely
as they are in the case of number of remiges in the ostrich wing,
but there is no indication that the changes are exclusively in a
minus direction, as Duerden assumes them to be in the ostrich.

e has observed variation in the number of plumes on the ostrich
wing ranging from 33 to 42. He assumes that the variation can
probably be carried below 33 by selection, through cumulation of
loss variations by dropping out of factors, but that variation in
the opposite direction is not to be expected because 42 is the
present maximum and factors for a higher number having once
been lost can not be recovered. Of course, the thing to do in order
to test the validity of this view is to give it an experimental trial,

1 Duerden, J. E., ‘‘The Germ Plasm of the Ostrich,’? Amer. NAT., 53,
p. 312

555

556 THE AMERICAN NATURALIST (Vou. LIL

and this, no doubt, Duerden is already doing. If the 42-plumed
cock has descendants with a higher plume number than 42, the
theory will have been disproved, which would undoubtly be
highly pleasing to Duerden because it would give him a more
hopeful basis for economic work. Now my own experimental
work with loss-variations leads me strongly to hold the more
hopeful view, that genetic changes are plus as well as minus, even
in the case of structures which are in course of phylogenetic de-
generation.

The degenerating lateral digits of the guinea-pig’s foot? pre-
sent a case parallel with those of the degenerating wing and the
degenerating fourth toe of the ostrich. The guinea-pig, like all
wild species of the genus, Cavia, has lost altogether the first of the
five typical digits, and has lost the fifth digit from its hind foot,
but not from the front foot. Some years ago I discovered a
guinea-pig which had an imperfectly developed fifth digit on one
hind foot. Neither of its parents had a fifth digit on either hind
foot. This fact ‘alone shows the possibility of plus fluctuation in
a degenerate organ. The polydactylous individual, a male, was
mated both with related and with unrelated females. By the
former, he had 13 polydactylous and 32 normal individuals; by
the latter he had 2 polydactylous and 30 normal individuals.
This result showed that normal females related to the polydactyl
male, even though themselves normal, transmitted a factor or
factors favorable to the production of the fifth toe, since more of
their offspring were polydactyl than of the offspring of ordinary
females, when both sorts were mated to the same polydactyl male.
Breeding the polydactyl offspring together and continuing the
race by selecting these individuals which had the best developed
toes (purely somatic selection), a race was secured within four
generations which produced regularly 90 to 100 per cent. of
polydactylous young. The race was continued for several years
and showed no signs during this period of returning deterioration.

In this case we have an example of plus fluctuation in a char-
acter supposed to have been completely lost from the genus, Cavia,
yet which, having shown itself sporadically and feebly in a single
individual, was recovered and fully established as a racial char-
acter by the practise of inbreeding and selection on a purely
somatic basis.

The first digit has, so far as I know, never been observed to

2 Cas tle, W. E., ‘<The Origin of a oboe ha eso Race of Guinea-pigs,’’
Publication No, 49, Carnegie Inst., Washington, 1906.

No. 629] SHORTER ARTICLES AND CORRESPONDENCE 557

occur in the genus, Cavia, except in the case of a single individual
born in one of our experiments. As this individual was still-
born we had no chance to experiment further in the case, but the
occurrence shows that degenerating characters are not of neces-
sity lost for ‘all time when they have ceased to have somatic ex-
pression in the race. I am therefore hopeful that Duerden will
live to see not only other 42-plumed ostriches but also those which
are 45-plumed or possibly even better, if selection for high num-
ber of plumes and inbreeding are persistently practised.

One point is worth noticing, which Duerden does not especially
emphasize, though it is highly suggestive. He notes the advanced
state of degeneration of the ostrich foot (presumably through
irrecoverable loss of factors) as seen in the complete disappear-
ance of digits 1, 2 and 5, and the greatly reduced size of digit 4,
which leaves the ostrich with practically a single functional toe
(digit 3), this being among birds an unparalleled amount of
digital reduction. He concludes ‘‘Should the loss of plumage
continue to a much further degree and marked degenerative
changes be set up in the big middle toe, Larabee selection may then
be expected to bring about extinction.’’ This, it seems to me, is
a needlessly gloomy view of the case. The fact that the middle
toe is ‘‘big’’ contradicts the idea that it will soon degenerate as.
the other digits have done. If evolution occurred only by loss
and never by gain, the middle toe could never have grown ‘‘big.”’
But in reality it has probably grown as the other digits have dis-
appeared. If so, factors must have been added to the genetic
complex, or plus factorial changes must have occurred by some
other means. Reduction in number of digits does not necessarily
mean degeneration. Note the parallel evolution of the horse.
Does any one consider it degenerate? Yet in the horse digital
reduction has gone even farther than in the ostrich and for a like
reason, increasing perfection of a cursorial type, for which one
good toe is better than three or five ordinary toes. Increase in
body size has occurred in both horse and ostrich concurrently
with digital reduction. It too has doubtless improved the eur-
sorial type, increasing its swiftness. Thus in the horse and in the
ostrich we have the culmination of cursorial types among mam-
mals and birds respectively. Each is highly specialized, but not
on that account degenerate or verging on extinction. Extinction
will come for each when man says the word but not sooner, so far
as we can foresee. Great specialization or great phylogenetic age
does not of necessity mean early extinction, if we may judge by

558 THE AMERICAN NATURALIST (Vou. LIII

the geological history of brachiopods, echinoderms and mollusks.
If a suitable environment continues, the specialized organism may
continue indefinitely. The idea that genetic variation occurs only

in one direction and is irreversible is widespread, but needs sub-
stantiation before we accept it into a category of fixed ideas. The
world indeed may wait long to see again a four-toed horse, but
the reason probably is that we already have a better type in the
one-toed horse, which replaced the former because it was better,
not because it was degenerate. If selection, natural or artificial,
saw at the present time a distinct advantage in a polydactylous
horse, it is quite possible that the type might once again be pro-
duced. The animal breeder would ask only such a start as was
seen in Cesar’s three-toed steed, to produce a race of polydactyl
horses. W. E. CASTLE.

BUSSEY INSTITUTION,
Forest HILLS—BosToN, Mass., October 1, 1919

ANOMALOUS RATIOS IN A FAMILY OF YELLOW MICE
SUGGESTING LINKAGE BETWEEN THE GENES
FOR YELLOW AND FOR BLACK

DuRrING the course of an experiment involving the breeding of
yellow and non-yellow varieties of mice certain anomalous ratios
were produced by a family of yellow mice. Since an explanation
of these facts brings out considerations regarding yellows which
have not been treated in the literature of the subject, it seems
well to put the case on record.

The peculiar family originated in a cross of black-and-tan (a |
very dark form of yellow) with brown. F, consisted of blacks
and yellows. The blacks when tested proved to be heterozygous
for brown and showed in their subsequent generations no pecul-
iarities of inheritance. The F, yéllows should theoretically have
been heterozygous for both black and brown for, .

Let oo NI parent (yellow carrying black)
and yybb = brown pare

Then F, should atau gr yellows, YybB, and binei, yyBb.

These F, yellows were back-crossed to pure browns.

The progeny distribution to be expected would be as follows:

The F, yellow parent, YybB, would form gametes, YB, Yb,
yB and yb,

The brown parent, yybb, would form only one type of gamete,
viz., yb. The expected zygotic combinations would be

The yellow young obtained from this back-cross should be of
two genotypes, YyBb (carrying both black and brown) and

No. 629] SHORTER ARTICLES AND CORRESPONDENCE 559

(1)

| YyBb, Yybb | yyBb yybb
DOLE TAMAR ong Ses ce Nc cua es | 2 yellows | 1 black 1 brown
Percentage expected ................! | 2 25
Percentage observed j. orio secre | 53.6 | 28.6 27-7.
Namper Observed). a o cs | 88 | 47 29

Yybb (carrying brown only). Two were selected for breeding
to determine in which genetic class each belonged. If both mice
were YyBb we should expect

2 YyBB 1 yyBB 1 yybb
2 Yybb
(2) 4 YybB 2 yyBb
8 yellow 3 black 1 brown
If both mice were Yybb, we should expect
1 yybb
(3) 2 yellow 1 brown
If one mouse were YyBb and the other Yybb, we should expect
i : 1 yyBB 1 yybb
(4) 2 Yybb ;
4 yellow 1 black 1 brown

The actual figures from this mating were (616 < 766) 14
yellow, 4 brown. This result resembles most closely that to be
expected if both parents were Yybb. If such were the case, all
of the yellows should carry brown only, never black. To test
them, two of the 14 yellows were mated with each other
(2160 xX 2162). They produced 18 yellow, 1 black and 10 brown
young. To account for the black young, we must suppose one
or both parents (2160 and 2162) to have been heterozygous for
black and hence that one of the yellow grandparents (616 or
766) carried black. Black young should have resulted from
their mating with each other but failed to do so. Black young
were also deficient in the mating of their descendants, 2160 and
2162. The ratio observed among the progeny of 2160 and 2162
indicates that one of them carried black rather than that both
of them did, for the expectation if both parents carried black, as
in (2), fits even more poorly than the expectation if only one
parent carried black, as in (4). The deficiency of black young
in the mating of 2162 and 2160 is shown by the percentage of
various young observed and expected:

Yellow Black Brown
Per Cent. expected........... 66.6 16.6 16.6
Por Gent. observed sis- eces 62.0 3a 34.5

560 THE AMERICAN NATURALIST [Vou. LIIE

Of the yellows resulting from this mating six when tested
proved to be heterozygous for black as well as brown, while four
carried brown only; and of the yellow young resulting from
these tests four were shown to carry black and brown, while
three carried brown only. Equality of yellows carrying both
black and brown and of yellow carrying brown only was ex-
pected in contrast with the observed ratio of 10 carrying black
and brown to 7 carrying brown only.

CONCLUSIONS

In the case reported the occurrence of black and brown reces-
sives out of crosses between yellows carrying both black and
brown is the reverse of that expected because brown (normally
recessive to black) has appeared in a frequency more than double
the expected, and black has appeared in a frequency less than
one third of the expected.

There are three theories which might explain these facts.
(1) Reversal of dominance resulting in the dominance of brown
over black. This may be discarded because the brown young in
this experiment were found not to carry black. (2) Selective
fertilization by means of which brown gametes united with brown
gametes in more than normal frequency. There is known at
present no mechanism for such a type of fertilization nor have
any cases of it been shown to occur. (3) Linkage of the genes
for yellow and black so that YB and yb gametes are formed more
often than yB and Yb gametes. Since in the above matings
black animals could only result from a combination of yB with
yB or yb the result of a linkage of Y and B would reduce the
number of blacks produced. This is substantially the result
obtained.

Of the three theories the last is favored because it affords a
satisfactory explanation of the observed facts in harmony with
other cases of linkage, and because it is more readily susceptible
_ of proof or disproof. It encounters the difficulty of positing a-
linkage between two genes, one of which (yellow) is either iden-
tical with or closely linked to a lethal gene and the other of which
has hitherto shown no evidence of being related to the lethal.

It is hoped that more data on the subject will be forthcoming
which will show whether the foregoing case is exceptional and due
to random sampling or whether the genes for yellow and for

black are commonly linked in the gametes of mice.
: | L. C. Dunn
BUSSEY INSTITUTION

No. 629] SHORTER ARTICLES AND CORRESPONDENCE 961

GENETICS AND EVOLUTION IN LEPTINOTARSA

In considering work of the kind presented by Professor Tower
in his latest partial report on Evolution in Leptinotarsa (Tower,
1918) it is necessary, in justice to the author, that we distin-
guish carefully between three different things: the actual experi-
mental work, the author’s interpretation of its results, and his
general speculations. Professor Tower had secured an unusually
favorable opportunity for attacking his problem, by a fortunate
selection of material. The ‘‘lineata’’ group of the genus Lepti-
notarsa comprises a large number of forms generally recognized
as ‘‘good species,” highly variable, crossing freely, often in-
habiting markedly different environmental complexes, and easily
bred in the laboratory. His work represents a prodigious amount
of painstaking labor, covering many years, under most favor-
able conditions. It will be a matter of regret to many that so
large a proportion of this volume has been devoted to very gen-
eral speculation, rather than a more complete presentation of
the large mass of interesting data which he has accumulated.

Considered as a contribution to genetics, there is ample con-
firmation, but nothing added, to what we already know, though
much that was unknown at the time this work was started. The
results of his extensive breeding experiments—including selec-
tion, intraspecific and interspecific crossing—show nothing, as
the author clearly points out, that is not perfectly Mendelian
and in accord with the factorial hypothesis. This portion of the
work is valuable, at least, as added refutation to the idea that
there is anything different involved in species crosses than in
intraspecific crosses.

One or two of the experiments, however, give results that seem
somewhat anomalous; and it is on these that the author has
attempted to build far-reaching hypotheses considerably at vari-
ance with those held by other geneticists. The most striking
ease, perhaps, is in connection with the crosses L. signaticollis X
diversa and L. signaticollis X undecemlineata. These crosses,
when made with material which had been bred in the laboratory
for several generations, gave normal Mendelian ratios (mono-
hybrid in the one case, trihybrid in the other). Certain strains
of signaticollis ‘‘fresh from nature’’ gave, however, in these
erosses, very complex and variable arrays. Crossed with diversa,
for example, there appeared in F,, besides the ordinary hetero-
zygote, breaking up normally in F,, a varying number (0-100
per cent.) of forms apparently pure signaticollis, which bred
true in all subsequent generations.

562 THE AMERICAN NATURALIST (Vou. LII

These experiments, to this point, had been previously pub-
lished by the author (Tower, 1910). Further work on this
apparently true-breeding race, however, has served to modify
the earlier conclusions as to its composition and behavior. Care-
ful measurements showed that this apparently pure signaticollis
race behaved normally as regards the form-index (relation of
width to length), a specifie character never dissociated, in other
experiments, from the conspicuous pattern difference involved.
Numerous experiments with this peculiarly behaving hybrid
failed, with one exception, to find any evidence of the diversa
pattern factor present. Individuals of this race having the
diversa (broad) form, crossed with one particular strain of sig-
naticollis, gave in F, four homozygous forms: pure signaticollis,
the peculiar true-breeding hybrid (diversa form with signati-
collis pattern), pure diversa, and a new form, with a different
pattern. These appeared in the ratio of 4:2:1:1 (actual num-
bers not given), with the corresponding array of heterozygotes.
` At first sight, all this seems easy to account for: one or both
of the parents in the original cross were heterozygous for one or
more factors. The actual results, however, can not be accounted
for, according to any known genetic principles, on the basis of
the facts given. There are many things at every stage of this
work, which we should like to know, about which we are told
nothing. The author states that the original parents were ‘‘not
heterozygous,’’ but no evidence of any genetic analysis is given.
In fact, the author declares more than once that he is not inter-
ested in ordinary Mendelian analysis, implying that it is unim-
portant. We do know that one, at least, of these parents was
“fresh from nature,’’ that is, genetically an unknown quantity.
The author himself shows very completely, in later chapters, that
these species, as found in nature, are genetically very hetero-
geneous. He also emphasizes that the results here obtained can
probably not be repeated with material from other locations or
even from the same location at other times. Again, we know
nothing of the genetic constitution of the signaticollis strain

in the ‘‘test reaction” where typical diversa is recovered ;

except that only this particular strain will give these Naolis.
The F, array in this case shows that there are at least two factor-
differences involved, while there is only one in the ordinary cross
L. signaticollis X diversa. Other crosses, which will immediately
suggest themselves as bearing on the problem, have not been
made or are, at least, not mentioned.

No. 629] SHORTER ARTICLES AND CORRESPONDENCE 563

Above all, it is to be regretted that in none of these experi-
ments have any definite pedigrees been given, nor the methods of
mating in the various stocks. The tables, moreover, are in most
eases merely summaries, the original and complete data not
being presented. It may well be that the full data and pedigrees
of all the author’s extensive experiments would be too bulky for
publication; but in view of the fact that on this particular
series only is based any claim of modification or addition to
genetic theory, the raw data and pedigrees might profitably have
been included, even to the possible exclusion of a part of the
discussion.

Under these circumstances, no attempt can be made at an
analysis by ordinary genetic conceptions. This whole experiment
illustrates perfectly the real basis of the familiar distinction
(implied also by the author in this volume) between results ob-
tained ‘‘in the laboratory’’ and ‘‘in nature.’’ The real differ-
ence, as well shown in this case, is merely one between working
with known materials under controlled conditions and uncon-
trolled operations with unknown things.

This particular experiment has been reviewed at length, be-
cause it is, apparently, solely upon it that the author bases a
large amount of rather far-fetched speculation on the ‘‘archi-
tecture of the germinal material,’’ and the influences of the ‘‘sur-
rounding medium’’ and the effects of crossing upon it.. His
interpretation of the results involves at least four independent,
and distinctly undemonstrated, assumptions: (1) that the ob-
served results are due, immediately, to a (reversible) change in
the relations of the genetic factors to each other; (2) that these
changes are the direct result of the crosses; (3) that these partic-
ular strains of signaticollis owe this peculiar behavior to the
direct effect of (cumulative) environmental influence on the rate
of ontogeny; and (4) that the results are profoundly and perma-
_nently modified by the environmental conditions surrounding
the parents of the original cross during the maturation of the
particular gametes involved.

It is needless to state that none of these hypotheses derives any
support from the experience of other geneticists; the only evi-
dence adduced by the author is that above described. The first
of these hypotheses is entirely inconsistent with all that we know
of the phenomena of linkage; if there is anything certain about
genetic factors, it is that they do not change their relations to
each other. The second hypothesis is evidently reminiscent of

564 THE AMERICAN NATURALIST (Vou. LHI

the pre-Mendelian conception that something new is produced by
the act of hybridization.

In his third hypothesis, the most characteristic and the most
involved of this group, there seems to be some confusion in the
mind of the author himself as to whether the environment has
produced its effect directly upon the gametes, or through its
modification of the rate of ontogeny. The latter, however, is
clearly implied in several places. But, beyond speaking of hypo-
thetical and rather vague ‘‘cytoplasmic determiners,’’ he makes
no attempt to even postulate a mechanism by which the rate of
ontogeny—in this case clearly, on the evidence, a somatic modi-
fication, directly dependent on temperature and moisture—can
effect genetic changes in the germ-plasm. In regard to the direct
effects (immediate or cumulative) of environmental conditions
upon the gametes, of which the author makes much throughout
his work, it can only be said that, however inherently probable
this hypothesis may be, there is no evidence for it in the present
work that will be acceptable to most geneticists. For the attempts
to distinguish between genetic and environmental factors are,
in this case, invalidated by an entirely inadequate analysis of
the former. The author assumes, throughout the work, that he
is dealing with materials of known genetic constitution ; but this
is obviously, on his own evidence, not the case. In the absence of
a more complete genetic analysis of the materials, there is no
valid reason to assume that there is anything involved beyond the
ordinary recombination of factors originally present.

Another anomalous result is found in the cross of L. decem-
lineata X diversa. These species cross with difficulty ; in only 2
cases out of 200 attempts was the cross successful. In both these
eases (decemlineata 2 X diversa g) the F, beetles showed all the
external characters of the mother, but the food choice and devel-
opment rate of the father. This hybrid bred true indefinitely,
showing no splitting up in F, or subsequent generations. This
kind of ‘‘stable hybrid’’ is familiar to us, from pre-Mendelian
literature, but has never withstood the test of rigorons Mendelian
and cytological investigation. Such a result can be produced
deliberately, where the gametic constitution of our materials has
been thoroughly analyzed, by means of ‘‘balanced lethals.’’
Should the present case be confirmed, and the numerous sources
of error eliminated, this explanation will undoubtedly be found
to cover it also. The possibilities of contamination or partheno-
genesis should, however, not be overlooked in such cases.

No. 629] SHORTER ARTICLES AND CORRESPONDENCE 565

In connection with his selection experiments, the author has
unfortunately adopted the theory of ‘‘germinal variations’’ (i. e.,
modifiability of the genes) since abandoned by its principal pro-
ponent. Tower’s evidence, like all the other evidence for this
theory, is based on a very evident confusion of biotypes with
phenotypes, and an apparent failure to recognize the vital im-
portance of rigid brother-sister mating in such experiments. In
the same experiments, the author finds a very complete refutation
(in this case, at least) of his own favorite theory of the direct
influence of the environment. In contrast to these indecisive or
negative results, there is at least one very beautiful and unques-
tionable demonstration of the effect of modifying factors (here
called ‘‘ genetic impurities’’).* Of the three methods by which he
attempted to limit the range of variation in the form-index of a
certain “‘biotype’’ (as he calls it) of L. multiteniata—namely,
control of the environmental complex, ordinary selection of ex-
tremes, and selection of an obvious modifying factor—only the
latter gave definite and permanent results. When certain indi-
viduals, showing a peculiar variation in pronotal pattern, were
selected and bred for this character, there was an immediate and
permanent reduction of the form-index variation. to less than
half its former range.

The most interesting and startling contribution made in the
author’s earlier report (Tower, 1906)—the production of muta-
tions in L. decemlineata by temperature and humidity, combined
with selection—is not again referred to in the present paper. In
regard to the supposed effects of these external conditions in the
present work—granting, in the absence of the details of his tech-
nique; that the author has been more successful than any other
worker in the exact control of humidity and evaporation-rate,
there is no evidence, which will be satisfactory to most geneticists,
that these factors had anything to do with the results observed.

The results of the author’s extended observations and experi-
ments with these same materials ‘‘in nature,’’ while generally in-
conclusive, like all such experiments, are in complete harmony
with the general conception of the processes of evolution held by
the modern adherents of the factorial hypothesis. The picture
we get of the conditions in this group of organisms is one of com-
plex genetic constitution, differing greatly in different localities,
resolvable entirely into simple Mendelian factor-differences, a
probable selective effect of environmental complexes, with a`

566 THE AMERICAN NATURALIST [Vou. LIII

direct modifying effect of temperature and moisture actually
demonstrated only as regards the rate of ontogeny.

ith regard to the general philosophie speculations of the
author, which occupy a large portion of this volume, little need
be said, as this sort of thing is largely a matter of taste. The
author states very unmistakeably that his viewpoint is purely
mechanistic, but there is much that will scarcely be accepted as
such by most mechanists. For instance, his list (p. 6) of the
‘‘ Categories of Organic Characteristics’’—(1) ‘‘Specifie Prop-
erties or Qualities,” (2) ‘‘Attributes’’ and (3) ‘‘Conditions’’—
will be regarded with suspicion by those to whom metaphysics is.
the béte noir of biology. The same may be said of his aversion
to ‘‘particularistic’’ theories of heredity. His objections, like
most other recent ones, are based on a map posed identification of
the gene with a somatic ‘‘unit character.’’ Yet he is himself
guilty, to an extraordinary degree, of such a confusion in his

‘‘scheme of classification of the agents of the germ plasm”?
(p. 86), where his hypothetical factors are classified entirely ac-
cording to their apparent somatic effects. This scheme is also
dominated by the ‘‘organism as a whole’’ dogma—that the
‘‘hasie’’ characteristics of the race are transmitted through the
cytoplasm, only trivial characters through the chromosomes and
capable of dissociation from the ‘‘species complex.

On the whole, it may be said that Professor Tower has con-
vincingly demonstrated the truth of his fundamental premise,
that ‘‘the general philosophical conceptions from which we
interpret nature will largely determine the logical, philosophical,
and expermentat methods used in investigation and the hypoth-
eses created.’ C. R. PLUNKETT

LITERATURE
Tower, Wak
1906. An Investigation of Evolution in Chrysomelid Beetles of the
Leptinotarsa. Carnegie Institution of Washington.

1910. The Determination of Dominance and the Modification of Be-
_ havior in Alternative (Mendelian) Inheritance by Conditions
Surrounding or Incident upon the Germ Cells at Fertilization,
Biol. Bull., 18.

1918. The Mochhaizei of Evolution in Leptinotarsa. beso Institu-
tion of Washington

IN DEX

NAMES OF CONTRIBUTORS ARE PRINTED IN SMALL CAPITALS,

ADAMS, CHARLES C., oe as a
Factor in Evolution
sare noe and ‘ pan Purpose-
ess,’’ FRANCIS B. SUMNER,
ALL A., Stud in sopr
Color’ Inhetiane in ar bacco, 79;
otiana pees Bid
Alternative ie serene
218; The Mendelian Behavior of
Aurea Character in a Cro 6 Be
tween Two Varieties of Ni boiin
“ia arened in
272
W. J. CROZIER,
nan ag and Adaptive Colora-
tion

Behavior and Assimilation, HENRY
D. H

Cercariae, Deseri
States, A Biological
ERNEST CARROLL FAUST,

CASTLE, W. E., Siamese, an 'AThinistie
Color Variation i in Cats, 265; Pie-
bald R an fallen "370;
Does Evolution occur exclusively
by loss of Genetic age k's 568

Cook À., rieties of
H elianthus tuberosus, Bh

Color Factors in Mice, C. C. LITTLE,

ibed, in the United

N, G. C., Evolution of Ar-

93; Resis
i onena Sea Water, 180
ence of the Shell Plates in
; ar gy EY
Onchidium an
Adaptive Galain 415

DUERDEN, J. E., Germ Plasm of the
Ostrich, 312
DUNN Bn C., Linkage between ae
for’ — and for Black
i

Egg-weight as a Criterion of Nu-
merical Production in the Domes-
tie Fowl, PHILIP HADLEY, 377

he Question of |

gg at Sparrow in Death Valley,
JOSEPH GRINNELL,
F oeitomi niai Reactions of Phryno-
, 33

Evolution, Migration m a Factor,
CHARLES C.

of Ar
thropods and Tits "Relatives, G.
y

, 282
by Loss of polaka Fac
Cas 568; and Gea ics
Tartini De PLUNKETT, 561

FAUST, ERNEST CARROLL, A Biolog-
o Surv rvey of Deseribed Cercariae

n the United Sta
Pando: Rosieinties ns t Concen-
rated Sea Water, W. J. CROZIER,

GALLASTEGUI, G. A and D. F.

JONES, Factor a ian in Maize
with reference to Linkage, 239.

and Evolution in aga
PLUNKETT, 561

Germ Piaam of the Ostrich, I E
agit 12

GLEA HL A., Variability in

Piat oib in Vernonia mis-

The English

E. e
Suc ing-fish for catching Fish and
Turtles 289, 446, 515

HADLEY, PHILIP, Egg-weight as a
Crite rion of Numerical Production
in the Domestic Fowl, 377

Heterophylly in Water

NES ARBER, 272
Henry D., Jr., Behavior
i 06

Plants,

LORANDE

tion to the Knowledge of, }
F "Ro ERTS

T F. ROBERTS

567

in

568

of Spiro-

ae amon g Kapas
EDGAR TRANSEAU,

NELSON
109”

IBSEN, HEMAN L., Synthetic Pink-
eyed Self White kieres Pigs, 120

Inheritance of Te n Oat

Hybrids, aE: W " Blos-
A. AL

ker Color i 5 Tobacco, i. ee
White Spotting and |
P. |

‘Other oie Peach.
W. WHITING, 473

Helianthus Ae miel Varieties of,
j A. COCKERELL, 188

in Cats,

JONES, D. F., and GALLASTEGUI, G
Biss Facto r Relations in rege
Reference to Linkage,

ag oe hogar Peers ee a Evolu-
C; , 561
Link age pekas en ae ‘Gait for Yel-
low and for Black in Mice, L. C.
DUNN,
LiTTLE, C. C., The Fate of Individ-
uals homozygous ee Certain Color
a tors in Mice
and fectieies: G.
Inheritance of Hull-lessness in oat
Hybrids,
Lunules, r A R and Reforma-
`~ tion in Mellita, W. J. CROZIER, 93

McRosrTIE, G. P., and LOVE wi
Inheritance of “Hull. Citi in
H: 35

Maize, Crosses, an oie Mg on,
HERBERT F. Factor
Relations in Maize w in Heference
to Linkage, D. F. JONES an A
GALLASTEGUI, 239; Morpholo,
Basis of some i | Wor
Spee _ PAUL WEATHERW
ea
ween Two

Varleties of Nicotiana sheet H.
A. ALLARD, 234

NABOURS, ROBERT K., Parthenogene-
sis ver

l and Crossing-o in Grouse
“Leeds Apotettix, 131
Nicotiana tabacum, and
its Alternative H.
A. ALLARD, 218
Onchidium dium Question of

and the
Adhie Coloration,
ZIERA

n ATE
d Lest B. AREY, pa

THE AMERICAN NATURALIST

ith |

heal |
k i

AX, 3
ge Behavior of Aurea Char-

WOODRUFF,

[ Vou. LIII

PARKER, G. H., Effects of the el
ter of 1917-191 8 on the Oce
rence of Sagartis Luciae Verrill,

cha tas and Crossing- ae
Gro moe — ust Apotett

Man T K. 131

Piebald "Rats eN ‘Galerkin: W. B
CASTLE, 370

PLUNKETT, C. R., Genetics and Evo-
lution in Leptinotarsa, 561

RıLey, E. é Curtis, Habitat Re-
sponses the Large ater-
strider, Gerris remigis Say, 394,

483
ROBERTS; HERBERT F.,

ts, s Contribu-
tion to is p tee aa of Hybridi-
zation,

Sagartis Luciae Verrill, the Effects
e Winter of 1917-1918 on

the Occurrence of, G. H. PARKER,
Shell Plates, pty of, in
Chiton, W. J. Croz
ee "an Albinistie Color Varia-

n Cats, W. E. CASTLE, 265
Bucking’ -fish, w Use of, oe ges

Purposefulness,’’ 193, 338

Maasen and avon Z., 282
TRANSEAU, EDGAR N, Hybrids
among y Fone oe of sort Aa 109
Vernonia missurica Raf., Verisbility
in Flower-number in, H. A. GLEA

’

Water- strider, Habitat Responses of,
C€. E. Curtis RILEY,

| WEATHERWAX, PAUL, ‘Merb

Basis o of some Experimental
A. O, Environmental eae:
tions of Phrynosoma 33
Wuirine, P. W. Inheritance of
White Spottin g and Other Color
a

Micrographia, 247