AM

\c\

LAND SNAILS FROM HAWAII,

CHRISTMAS ISLAND,

AND SAMOA

Bv

HENRY A. PILSBRY, C. MONTAGUE COOKE, JR.,
AND MARIE C. NEAL

Bernige p. Bishop Museum

BtJLLETIN 47

HoNOLuiu, Hawaii

PtjBUSHEP BY THE MuSEUM

/o 3

LAND SNAILS FROM HAWAII,

CHRISTMAS ISLAND,

AND SAMOA

BY

HENRY A. PILSBRY, C. MONTAGUE COOKE, JR.,
AND MARIE C. NEAL

Bernice p. Bishop Museum
Bulletin 47

Honolulu, Hawaii

Published by ths Museum

1928

A^.

101

,««- t

, /d 5 1

CONTENTS

Pag«
Georissa, a land snail genus new to the Hawaiian islands, by Henry A. Pilsbry 3

Food habits of Partula zebrina Gould, by C. Montague Cooke, Jr 5

Introduction 5

Food of Partula 7

Anatomy 11

Summary 12

Three Endodonta from Oahu, by C. Montague Cooke, Jr 13

Introduction 13

Habits 14

Endodonta lamellosa 14

Shell characters 16

Soft anatomy 19

Conclusion 21

Endodonta marsupialis 21

Shell characters 22

Soft anatomy 22

Conclusion 24

Endodonta fricki 24

Shell characters 24

Soft anatomy 25

Conclusion 27

Bibliography 27

Distribution and anatomy of Pupoidopsis hawaiiensis, by C. Montague Cooke, Jr.,
and Marie C. Neal 28

Distribution 28

Anatomy 29

General 29

Jaw and teeth 30

Kidney 31

Genitalia 31

Summary ^^

Anatomical studies of Achatinellidae, by Marie C. Neal 34

Comparison of representative species of the family 34

Development of genitalia in Achatinella viridans 44

ILLUSTRATIONS

Figure 1. Three new species of Georissa 4

2. Mouth parts of Partula zebrina and shells swallowed 11

3. Anatomy of Endodonta lamellosa 18

4. Free muscles and teeth of Endodonta lamellosa 19

5. Reproductive organs of Endodonta lamellosa 20

6. Reproductive organs and teeth of Endodonta marsupialis 23

7. Reproductive organs and tooth of Endodonta fricki 26

8. Expanded animal, pallial organs, and alimentary tract of Pupoidopsis

hawaiiensis 28

9. Reproductive organs of Pupoidopsis hawaiiensis 30

10. Reproductive organs, teeth, jaw, and position of tentacles of Pupoidopsis

. hawaiiensis 32

11. Pallial organs and alimentary tract of Partulina redfieldii, P. horneri,

P. virgulata, and P. mauiensis 35

12. Pallial organs and alimentary tract of Newcombia canaliculata and

Achatinella lehuiensis var. meineckei 36

13. Genitalia of Partulina redfieldii 36

14. Genitalia of Achatinella lehuiensis var. meineckei, Partulina virgulata,

and P. rufa 38

15. Genitalia of Partulina mauiensis, P, horneri, P. dubia, and Newcombia

cumingi 39

16. Nerve collar of Partulina redfieldii 42

17. Genitalia of Achatinella viridans 44

18. Genitalia of Achatinella viridans 45

19. Genitalia of Achatinella viridans 46

20. Genitalia of Achatinella viridans 47

21. Genitalia of Achatinella viridans 47

22. Genitalia of Achatinella viridans 48

Land Snails from Hawaii, Christmas
Island, and Samoa

By
He;nry a. Pilsbry, C. Montague: Cooke:, Jr., and Marie: C. Neai.

GEORISSA, A LAND SNAIL GENUS NEW TO THE
HAWAIIAN ISLANDS

By he:nry a. pii^sbry

In the course of a trip to Kauai, Marie C. Neal, Assistant Malacologist,
Bernice P. Bishop Museum, had the good fortune to collect specimens of
three species of Georissa, forming the subject of this paper.

The special interest of this discovery lies in the fact that it adds a
genus and a family new to the Hawaiian fauna. It is, moreover, a group
of purely terrestrial station, not a seashore or atoll dweller such as Trun-
catella; therefore the less likely to be transported by natural rafts or other
means of oversea drift. The nearest known neighbors of these Kauaian
Georissa are in Guam and in the Society Islands. The species here described
appear to be more closely related to those of the Society Islands than to
species of Guam. On the western borders of the Pacific, the genus Georissa
is known from Japan, China, the Philippines, and southward to tropical
Queensland. Eastward it is spread through the Cook and Society islands
to the Marquesas.

All of the species were taken on a ridge between Hanalei and Wailua,
Kauai, on the Pole Line Trail between Kualapa and the summit camp,
elevation about 2,000 feet, in moss collected on fallen logs. The moss was
subsequently dried, and the shells were found in the finest si f tings, the
only other shells occurring with them being young specimens of Tornatellina
and Leptachatina and a single specimen of an undescribed species of Punctum,

Georissa cookei, new species (figure 1, a).

The shell is conic, orange-cinnamon colored, composed of 3^ strongly convex
whorls. The first whorl appears to be smooth or very minutely punctate; the last
two whorls are encircled with low spiral cords about as wide as their intervals. At

Bernice P. Bishop Museum — Bulletin 47

the beginning of the last whorl are about 5 or 6 cords between lip and suture; on
the base the spirals are smaller and weaker. The oblique aperture is somewhat wider
than semicircular. Columellar margin straight, heavily calloused.

Length i mm., diam. o.8 mm., height of aperture, 0.5 mm.

Figure 1. — Georissa: a, Georissa cookei, new species; b, Georissa neali, new
species ; c, Georissa kauaiensis, new species.

Type, 131432, Academy of Natural Sciences of Philadelphia; paratype,

59188, Bishop Museum.

Georissa neali, new species (figure 1, &).

The shell is decidedly shorter than G. cookei, of 2^2 strongly convex whorls. The
embryonic shell is inflated, hemispherical, deeply and closely striate across (Trivia-like,
as seen from above). Subsequent whorls have a few low, wide, somewhat uneven spiral
cords. The aperture is strongly oblique, semicircular, the straight columellar margin
heavily calloused. Color white, but probably bleached.

Length 0.85 mm., diam. 0.75 mm.

Type, 131433, Academy of Natural Sciences of Philadelphia; paratype,

59189, Bishop Museum.

Georissa kauaiensis, new species (figure 1, c).

The shell is light pinkish cinnamon, somewhat more globose than G. neali, with
the first whorl of more regular shape, not striate, but showing a faint, close microscopic
punctation; 2^ whorls, the last having faint traces of spiral sculpture.

Length 0.9 mm., diam. 0.8 mm.

Type, 131431, Academy of Natural Sciences of Philadelphia; paratype,
59187, Bishop Museum.

The type of this species is evidently not adult, but it differs from its
fellows so much in sculpture that it must represent another species. The
embryonic whorl appears to be larger than in G, cookei, and subsequent
whorls are far less strongly sculptured.

Very few specimens of these tiny shells have been examined, and their
limits of variation are unknown. They are the smallest species yet known
in the genus.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 5

FOOD HABITS OF PARTULA ZEBRINA GOULDi

BY C. MONTAGUE COOKK, JR.

INTRODUCTION

The family Partulidae contains but the single genus Partida. This
family is confined to high islands of the Pacific and has no near relatives
on any of the continents bordering this ocean. The eastern limits are
Rapa and the Marquesas islands, the western the Pelew, Talauer, and
Admiralty islands. It does not occur in the Hawaiian islands to the north,
and its southernmost representative is on the island of Rapa.

Most of the literature on the genus Partula is taxonomic. While little
is devoted to distribution, still less has been written on the habits of the
snails. Crampton is the only writer who has closely studied the habits of
the genus. 2 Other than that certain species are arboreal or terrestrial I can
find nothing else that has been published. Crampton found that

... the greater number of species in Tahiti, in the Society Islands, and elsewhere,
whose representatives are arboreal during the day time . . . [remain] sealed up on the
under sides of leaves less than 10 to 15 feet from the ground; at night, however, they
resume activity and crawl to the earth to feed upon decaying vegetation. . . . Under
normal conditions . . . Partulae resume activity at night and become quiet in bright
light, in practically 95 per cent of the cases as observed in the laboratory, and in ex-
periments in the field where snails were placed on suitable plants at the coast for
determination of this point. But if the atmosphere is very dry, they are apt to remain
sealed up all night, while, on the other hand, if the sky becomes clouded and the
light is reduced, they will crawl about even in the day time. When rain falls and
humidity increases greatly, they move about with celerity, crawl to the ground, and
resume feeding.

As far as I can learn the only observations in regard to their food
have been made by the same author. To quote again:

The character of the food of Partulae is a matter of considerable interest, in view
of the theoretical possibility that differences in the vegetation of separated valleys
might be adduced to account for the diverse characteristics of the snails in such dif-
ferent localities. When the facts are examined, however, this possibility is ruled out.
Confining our attention to the low arboreal species, and disregarding those of the
ground and of the highest tree-tops, we have several independent series of data that
justify the statement given. In the first place, it is clear that the snails do not feed
on the living tissue of the plants upon which they are found, because the organisms
may be present in numbers on a perfect unmutilated leaf, and because no leaves occur
that display alterations saving such as are made by insect larvae. Secondly, the animals
go to the ground to feed, as described in the foregoing paragraphs. Thirdly, the
stomach-contents of active and newly killed snails comprise no fresh vegetable material,

» Read at the second annual meeting of the Hawaiian Academy of Science. ^^^^ J. 1927.
» Crampton, H. E., Studies on the variation, distribution, and evolution of the genus I'artula.
Carnegie Inst. Washington, Pub. 228, pp. 24-26, 1916.

6 Bernice P. Bishop Museum — Bulletin 47

but only the rasped fragments of decaying woody and fleshy plant tissues. Again,
animals that were feeding under observation in nature avoided the living parts of
plants and restricted themselves to decaying leaves and wood. Finally, the specimens
that were brought back to the laboratory ate moist dead leaves with avidity ; only
when very hungry did they eat the whiter and firmer portions of lettuce and cabbage
leaves. In transporting them from the field, they consumed great quantities of wet
paper, which to all intents and purposes, is wood pulp. Dall has suggested that in
the case of Achatinellidae decaying vegetation is eaten for the sake of the fungi that
grow in and upon it. Whether or not the same is true for Partulae, the fact remains
that the nurse-plants upon which Partulae are found during the day do not seem to
have any effect upon the specific or varietal characters displayed by these animals.
We would scarcely expect that this would be so, when representatives of three distinct
species may be sealed up on a single leaf of caladium, plantain, or Dracaena,

Last year in company v^ith Mr. A. F. Judd and Mr. Theodore T. Dranga,
I spent six weeks in American Samoa. On the Island of Tutuila three
species of Partula were taken. These three species were found associated
in diflferent proportions, in some localities, but in many localities only the
most common species, Partula zehrina Gould, was found. The habitats of
two of the species, Partula conica Gould, and Partula abbreviata Mousson,
were also slightly different from those of P. sebrina. The first two species
were usually taken on the trunks and branches of trees and shrubs but
sometimes on the leaves, while P. zebrina was taken under almost all
conditions, usually on the leaves of shrubs and trees, sometimes on trunks
and branches of trees. A considerable per cent were also found on and
under dead leaves on the ground and a very few on stones. No obser-
vations were made by any of us as to whether they were nocturnal in their
habits, but it was very noticeable that a large number of specimens were
active and crawling during rain and shortly after.

Nearly all our material was expanded by drowning and brought back to
Honolulu in alcohol. During the early stages of the process of extracting
animals from shells. Miss Neal found a specimen of Aphanoconia in the
body of a young specimen of P. zebrina, which led to a tentative examina-
tion of every specimen of Partula that was cleaned in the laboratory. All
animals in the stomachs of which shells were found were isolated for fur-
ther study, and some of them served as the material on which this paper is
based. I have used only a portion of the data collected and have examined
critically the stomachs of about 200 specimens.

The catalog numbers of the specimens referred to in this paper are as
follows: AmouH, 84173-84185, 84228, 84234, and 84238-84243; Pago Pago,
83779-83789 and 83821-83831; Fagasa (arboreal), 83919-83933 and 83937-
83942; Fagasa (terrestrial), 83970-83975; Mount Tau, 84457; Logotala,
84783.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa

FOOD OF PARTULA

Judging from the type of the teeth of Partula, the genus is herbivorous.
As shown by Crampton, the stomach contents of Tahitian species is made
up of decayed vegetable matter, which was found also in some of the
Samoan snails. To determine this point, the stomachs of 50 P. sebrina
were examined. Of these, nine were empty or nearly so, and one or two
were found to contain matter that could be classed strictly as decayed vege-
table matter, as it was very black. In most specimens the vegetable material
consists of very fine brownish particles, which appeared to be made up of
rather newly dead vegetable matter. In addition to this material, 21 speci-
mens had the stomachs crowded with or partly full of a compact white
mass. Dr. Brown has made a careful microscopic study of this material
and has come to the conclusion that it is the undigested remains of vegetable
matter. In the stomach of one specimen, in addition to the white mass
already mentioned. Dr. Brown was able to recognize partly digested, newly
dead leaf cells and fibers, diatoms, Protococcus, fern sori, bacteria, pollen
grains, etc. Of the 50 specimens examined, only three contained the remains
of shell animals. As the shells had been completely dissolved and the animals
crowded in with vegetable matter, none of the genera was recognizable.

Of the three species of Partula occurring on the Island of Tutuila, only
P. zehrina was found to have acquired the habit of swallowing other land
shells. Of the many hundreds of specimens of P. conica and P. abbreviata
that were tentatively examined, not a single stomach was found to contain
a shell. This is probably due to the relatively thicker shell of P. zebrina,
and, as I will try to show later, shells are swallowed merely to procure lime
for the formation of new shell material.

The frequency of this habit is illustrated by Table 1, for which I have
selected specimens from five somewhat separated colonies.

TABI.E: 1. I^RKQUE:nCY Ot THE SHElvI.-SWAI.I.OWING HABIT IN PARTUI^A

ZKBRINA

Colony

Amouli

Pago Pago

Fagasa (arb.)
Fagasa (ter.)
Mount Tau ...

Logotala

Totals

ADULT

u

en

1

CO

X!

OS

5

^^^

138

22

116

16

72

22

50

30

120

21

99

17

6

2

4

33

50

13

37

26

137

9

128

6

523

89

434

17

YOUNG

W5

u

^

(U

<u

3

X!

1^

to*

:^

.t^

.tn "^

^

^

27

14

13

52

52

30

22

58

43

17

26

40

9

5

4

55

20

11

9

55

18

9

9

50

169

86

83

51

8

Bernice P. Bishop Museum — Bulletin 47

The table is based on nearly 700 specimens. It shows that the occurrence
of this habit in juveniles is almost exactly three times as frequent as in
adults. This is undoubtedly due to the great need for lime in young growing
shells.

The species of shells and the number of individuals swallowed was tabu-
lated from 43 specimens of adult and 40 specimens of juvenile stomachs. A
summary of the results is seen in Tables 2 and 3.

TABIvE 2. NUMBER O^ SHEl^IyS SWAI,I,OWKD BY PARTUI.A ZKBRINA.

ADUI.T

YOUNG

Colony

Number

in
stomach

1

2

~5
6
3
3

3

4

5

6

7

Number

in
stomach

1

4
8

7

2

3
2
4

~9

3
6
1

T

4

1
2

3

5
1

1

6

7

Amoiili

Pago Pago

Fagasa

Mount Tau ....

10
10
10
13

3
4
7
9

2

1

16
12
12

1

Totals

43

23

17

2

1

40

19

1

From Table 2 it is seen that 43 adult specimens had swallowed 70 shells
with an average of nearly 1.7 shells to each individual, while 40 juveniles
had swallowed 82 shells with an average of just over two shells for a meal.
Of the 83 specimens, 42 had swallowed 1 shell, 26 had swallowed 2, 9 had
swallowed 3, 3 had swallowed 4, 1 had swallowed 5, and 2 had swallowed
7 specimens.

As shown in Table 3, of the 152 shells that had served as meals, 140
were terrestrial in habit, only 12 arboreal. Only 5 of the 13 terrestrial
genera found in Tutuila seemed to have been swallowed by P. sebrina.
Whether the other terrestrial genera (Trochonanina, Patula, Succinea,
Gastrocopta, Gulella, Tornatellina, Ostodes, Assiminea) are distasteful I do
not know; but two at least (Ostodes and Assiminea) would rank among the
five most abundant genera in point of individual numbers. No specimen
of the very abundant arboreal genera Trochonanina nor Trochomorpha nor
Partula was found in any of the stomachs examined. It is probable that
some of these species, especially the arboreal, are too active to be caught.

Most of the shells which had been swallowed were alive when they
reached the stomachs of their captors. This was clearly shown by the large
number of specimens with the animal fully or partly expanded. Several
animals were as fully expanded as in freshly drowned specimens. None
of these shells had been attacked by the digestive fluid.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 9

table: 3. GE:NERA of SHE:i.LS SWALI^OWED by PARTUr^A ZKERINA.

Amouli
Pago Pago
Fagasa

Total

A D U 1. T

Colony

e

CO

1

2

1

►5^

i

11

1

g

3

r

1

w

m

f^

m

3

6

5

19

h^t

<

Amouli

5

19

Pago Pago

1

2

2

3

8

16

13

3

Fagasa

2

10

1

13

11

2

Mount Tau

3

7

2

3
20

3
9

2

23

22
70

19

3

Total

1

2

8

62

8(11%)

YOUNG

24

2

13

5

44

44

3

8

3

6

20

17

1

11

3

3

18

17

1

3

43

5

16

14

82

78

2

19

2

63

14

24

37

152

140

4 (5%)
12

Of the 41 specimens that had swallowed more than one specimen, 17
had shown some discrimination by confining themselves to a single genus,
22 had representatives of two genera, and only 2 had specimens of three
genera in their stomachs. (See Table 4.)

TABI,E 4. NUMBER OF GENERA IN STOMACHS OF PARTULA ZEBRINA

ADUI,T

Number of specimens
in stomach

One

genus

Two
genera

Three
genera

2
3

7

8
1

9
1
1

9

11

YOUNG

2

3

6

3

5

2

4

2

1

5

1

7

1

8

11

2

10 Bernice P. Bishop Museum — Bulletin 47

Slightly more than half the specimens {yy) that had been swallowed
belong to the genera Suhulina and Opeas. Although not indigenous to the
Samoan fauna, these genera would easily rank in abundance of individuals
among the first five. I consider that the high per cent of representation
is due to their abundance rather than that P. zebrina shows preference for
these genera.

Only four shells were found that had died before being swallowed. I
do not think that this shows conclusively that live shells are preferred, but
that due to conditions under which P. zebrina were feeding, namely, on
dead leaves, dead shells would scarcely occur.

All the snails that have conic or oblong shells, in which the shell is
somewhat longer than broad, were swallowed base first. I consider that
this is due to their habits of feeding. From the fact that feeding snails
have the mouth in contact with the surface on which they are crawling
they would naturally come in contact with the base of their prey rather
than with the apical portion.

Only a very few shells that had been swallowed were in perfect condition
when examined. A great majority had portions of the shell more or less
dissolved by some secretion. That the shells were swallowed primarily for
their lime was shown by the very slight effect that the secretions in the
stomach had upon the animals. About one-tenth of the swallowed shells had the
animal partly or fully expanded, showing that the digestive ferments were non-
irritating or the animals would have contracted into their shells. A con-
siderable number of individuals were noted in which the delicate surface
color markings of the animals were intact, though quite large portions of
the shell had been dissolved. All the lime of some shells had been dissolved,
with the animal and organic matter of the shell m such perfect condition
that not only the genus but the species of the individual could be recognized.

That these secretions did not penetrate into the animals was seen in
three or four specimens of Suhulina or Opeas that were pregnant with
eggs. In these, although the shells were considerably dissolved, the ^gg
shells were not dissolved or only dissolved in the portion that was in prox-
imity with the outer wall of the animal. One interesting specimen of
Subulina was found in which two tgg shells had been dissolved, but the
very delicate embryonic shells embedded in the albumen of the eggs were
not attacked.

Most of the stomachs that contained shells contained little or no vegetable
matter; but a few were crowded with vegetable material and had shells in
addition.

Pilshry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 1 1

ANATOMY

The mouth of an average specimen of Partula zehrina measures extern-
ally slightly less than 2 mm. in its longest diameter. Back of the mouth is
the buccal mass, which consists of a rather solid series of muscles for work-
ing the jaw and radula. Attached to the upper portion of the buccal mass
is the esophagus, a long, rather slender tube leading to the crop or stomach.
The esophagus is from .75 mm. to 1.25 mm. in diameter. Shells have been
taken from the stomach of a P. zehrina measuring 7 mm. in diameter. That
individuals swallowing shells must be put to a great deal of inconvenience
is to be assumed, since the diameter of a shell may be from seven to ten
times that of the esophagus. (See fig. 2, a, b, c, and ^.)

Figure 2. — Partula zehrina, showing shells swallowed and mouth : a, Partula
zehrina Gould, lower surface of the animal, showing mouth and fore end of foot ;
h, buccal mass of the same individual dissected out, showing esophagus, e. = eso-
phagus, p. r. = pharyngeal retractor; c, specimen of Aphanoconia fulgora Gould taken
from the stomach of the same individual, dashes represent where the shell was eroded;
d, another individual with the buccal mass extruded, e. = esophagus, 1. = lip, m. = mantle,
t. = tentacle, dashes represent the position of the buccal mass ; e, specimen of Opeas
taken from the stomach of a third individual.

(All figures drawn with the same magnification.)

12 Bernice P, Bishop Museum — Bulletin 47

The mechanics of this performance must be intensely interesting. Fortu-
nately a few specimens were found that offer a clue as to how this is accom-
plished. In these specimens the whole buccal mass and part of the esophagus
was extruded beyond the lips. (See fig. 2, d.) Probably a lime-hungry
P. zehrina coming in contact with its prey extrudes the mouth, which by
pressure enlarges and surrounds the victim. Then by contraction of the
mouth the shell is forced into the esophagus and thence into the stomach.

One difficulty has to be met before this is accomplished for the first
time. The central nervous system of a snail is made up of a number of
ganglia united by commissures to form a ring, through which the esophagus
passes. In P. conica and abbreviata these ganglia are united by relatively
short and thick commissures so that the ring fits rather closely to the
esophagus and is only slightly larger in diameter. In some individuals of
P. zebrina that had swallowed shells, this ring was enlarged to approximately
three times the diameter of that of a species of Partula with normal habits,
although the ring was not as large as the undissolved shells that had passed
through it a short time before. The ganglia of P. zebrina are united by
relatively much thinner commissures than those in species of Partula with nor-
mal habits. In one specimen of P. sebrina that contained only vegetable matter
in the stomach, the ganglion ring was more contracted but still fitted loosely
around the esophagus. Undoubtedly this ganglion ring is stretched to a large
extent when an individual swallows a relatively large snail, afterwards
contracting somewhat, but not regaining the size of a normal ganglion ring
of the genus. Evidently, therefore, not only the mouth and esophagus and
other alimentary parts of the snail are greatly stretched, but also the part
that functions as the brain of the animal is considerably altered.

SUMMARY

The species Partula zebrina, an herbivorous snail, has acquired the habit
of swallowing other species of snails. From the relative size of the victims
this must be accomplished with much inconvenience, if not pain, as not only
the alimentary organs but even the central nervous system are stretched
to many times their normal size. Since the secretions in the stomach
apparently attack only the calcareous material, leaving the animal matter
practically untouched, at least until some time after the shell is completely
dissolved, it is presumed that these shells are swallowed solely to procure
lime. This can be explained also by the relatively larger proportion of young
specimens showing this habit than adults, as the growing shell needs a larger
amount of lime.

Pilsbry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 13
THREE ENDODONTA FROM OAHU ^

BY C. MONTAGUE: COOKE^ JR.

INTRODUCTION

This paper is an attempt to identify and define the three species of
Endodonta (sensu stricto) that have been described from Oahu: E.lamellosa
Ferussac, B. marsupialis Pilsbry and Vanatta, E. fricki Pfeiflfer. Among
collections their exact determination is in doubt, and in attempts to clarify
the situation different authors have added somewhat to the confusion. The
exiict determination of Endodonta lamellosa I feel sure of, as the figure
given by Ferussac (2)* represents in its main characters shells from the top
of Konahuanui. Though the original specimens probably did not come from
this restricted colony, doubtless they were obtained at no very great distance
and at a considerably lower elevation.

Two paratype specimens of Endodonta marsupialis were kindly deposited
in the Bernice I*. Bishop Museum by Dr. Pilsbry, and I collected additional
typical material on the Pauoa slope of Mount Tantalus. Undoubtedly Dr.
Pilsbry's original material was collected in the immediate vicinity, as that
region is easily accessible to Honolulu and the additional material agrees
closely with the paratypes.

Of Endodonta fricki I feel fairly certain. In 1905 I compared three
specimens (Catalog No. 14291) with the supposed types in the British
Museum and identified them as this species. Unfortunately the comparison
w^s not as critical as I could wish^ According to my notes the only marked
difference is a proportionately higher spire in the original specimens. When
preparing this paper I noticed that Pfeiffer (7) mentions one important
character not found in any Oahuan species of the typical section, namely, a
margined suture. This character, if distinct, belongs to a quite different
group of species, which cannot be included in Pfeiffer's diagnoses and cer-
tainly does not approach his figure, which is of a typically Waianae species.
The number of whorls given by Pfeiffer does not agree with any of the
Waianae forms known to me.

The locality from which Pfeiffer's type specimens came can be deter-
mined only on comparison with a number of carefully collected specimens
representing the numerous geographical races.

' This paper was submitted to Dr. Henry A. Pilsbry for suggestions. In a letter dated Novem-
ber 3, 1927, he says: "I have reviewed the subject in the light of your researches and agree with
your results. ... At the time I wrotfe, only very few people appreciated how local those Hawaiian
species are; certainly I did not. I have gone over our shells and believe your conclusions as to
E. fricki are correct."

^ The numbers in parentheses refer to Bibliography, page 27.

14 Bernice P. Bishop Museum — Bulletin 47

HABITS

All the Hawaiian species of Endodonta (sensu stricto) are terrestrial
and have strictly sedentary habits. Colonies of some of the species found
only a few hundred yards apart and separated by some natural barrier, such
as a barren ridge or absence of forest growth, are easily recognized as
distinct geographical races. Some of these races are more circumscribed
in their known habitats than are the well-known geographical races of
Achatinella. A few species are more widely distributed and do not appear to
have given rise to distinct races.

In population and numbers, the two mountain ranges of Oahu offer a
decided contrast. On the longer Koolau Range living specimens have been
taken, as far as I know, in only five or six localities; though at one time,
when the island was more heavily and evenly forested, as evidenced by
fossil material, this genus was widely and abundantly distributed over the
whole lower region, down to and including parts of the coastal plain. On
the other hand, the shorter and drier Waianae Mountains, only a few miles
distant, have a very abundant and varied Endodonta faunula.

In the Koolau Range four of the five colonies have been located on the
higher, very damp mountain peaks at altitudes of 2,500 to 3,000 feet. The
only exception is a colony on the side of Mount Tantalus at an elevation
of about 1,700 feet. All the specimens from the Koolau Range were taken
on dead leaves and twigs. In the Waianae Mountains the representatives
of this genus do not seem to be so particular as to habitat. Almost any
suitable locality, as a mountain ravine more or less covered with forest
growth and fairly damp, and some unsuitable localities, dry and exposed,
may each be inhabited by a more or less well-populated colony of one species.
However, in some localities that are apparently suitable not a shell can be
found. In many colonies that occupy a small area the specimens are
abundant and by their great similarity give evidence of descent from a single
individual or at least a very few individuals and of close in-breeding. As
far as I am aware, no colony of Endodonta has been found in a talus slope
in the Koolau Range. Several isolated colonies have been found living in
talus slopes of the Waianae Mountains, some of them in dry and exposed
situations.

ENDODONTA LAMELLOSA

Endodonta lamellosa (Ferussac)

Helix (Helicodonta) lamellosa Ferussac: Prodrome, p. 35, no. 110, 1822 (nude
name) ; Hist. Nat. Moll., pi. 51A, fig. 3; Voy. Freycinet, Zool., p. 469, 1824.

Helix lamellosa Deshayes: Hist. Nat. Moll., vol. 1, p. 369 (French not Latin
description).

Endodonta lamellosa Albers : Die Heliceen, p. 89, 1850.

Pilsbry, Cooke, Neat — Snails from Hawaii, Christmas, Samoa 15

In the Prodrome, Ferussac described this species by name only, giving
the locaHty as "Islands of the South Sea." In the reports of Freycinet's
voyage, he republished this species and refers to the plate and figure in the
Histoire Naturelle des Mollusques, but gives "Port Jackson and the Islands
of the Great Ocean'' as the habitat. On the same page, under H, contorta,
he furnishes a short description of H. lamellosa, mentioning only that it is
"prettily spotted" and has "eight elevated lamellae obstructing its aperture."
He also states that the specimens are young and compares this species with
Helix alhella.

In 1845 PfeiflFer (4) described this species, undoubtedly from specimens
in his collection and quoted this description in all its essential characters
in two later publications (5,6). Deshayes (1) quotes Pfeiffer's Latin
description in the Histoire, but has rewritten a rather full description in
French. As this description differs in a number of essential characters from
the description by Pfeiffer, I feel quite certain that it is from the original
material which served as the model for the figure in the Histoire.

If Pfeiffer's description is based on the shell figured in the Conchylien-
Cabinet (5) these references to lamellosa should be dropped. Judging from
these figures, the specimen was close to if not identical with H. marsupialis
described by Pilsbry and Vanatta in 1905. The proportionately high spire,
flatter base, the narrow umbilicus and the prolongation of the aperture onto
the edge of the umbilicus, as shown in Pfeiffer's figure 9, lead me to this
conclusion, as these are a combination of characters that do not occur in
the species figured by Ferussac.

Tryon (11) quotes the more important characters of Pfeiffer's descrip-
tion and reproduces his figures.

From the rules of nomenclature Ferussac's figure in the Histoire has
precedence over Pfeiffer's description in Zeitschrift fiir Malakologie and
Pfeiffer's figure in the Conchylien-Cabinet.

Pilsbry 's figures of £. lamellosa (8) are undoubtedly drawn from speci-
mens referable to E. fricki. The form and proportional size of the umbil-
icus, the suffused brownish color of the base, and the two outer smaller and
more deeply seated basal laminae are a combination of characters of a
geographical race abundant in the western valleys of the Waianae Moun-
tains back of Mokuleia.

In 1905 Pilsbry and Vanatta (9) refigured lamellosa from specimens
reported from the island of Kauai and referred specimens from the Waianae
Mountains and also from West Maui to this species. The figured specimen
has all the characters of a shell from the Waianae Mountains. The rela-
tively wider spiral terrace extending into the umbilicus for about 1^
whorls, the proportional height of the spire, and the large size (9.5 mm. in
diameter, 63^ whorls) of the shell are a combination of characters at

i6 Bernice P. Bishop Museum — Bulletin 47

present only known in shells from the Waianae Mountains. Both of the
Kauaian species, of which I have examined a number of series, both recent
and fossil, are characterized by having at least the post-nuclear if not all
the whorls transversely costate or distinctly striate.

In 1871 Pease (3) united B. fricki and B. lamellosa. In this he was
followed by Sykes (10), who added Mount Kaala of the Waianae Moun-
tains and the mountains of Lanai to the habitats. But I see no justification
for uniting them.

Judging from Ferussac's figure of what was undoubtedly the type speci-
men, I am fairly certain that this specimen came from the slopes of Mount
Konahuanui, the highest peak of the Koolau Range. In the collection of
Bernice P. Bishop Museum are several small lots of this species from
this locality. They are not abundant; but specimens are found on and
under dead leaves and twigs at and near the top of the peak, within an area
of not more than 200 to 300 square yards. In only one other locality of
which I have any knowledge are shells approaching Ferussac's figures — ^the
top of Puu Kahuauli, about three miles distant from Konahuanui, at the
western head of Kalihi Valley. I have examined only two specimens from
this locality, one collected by Gouveia and the other by Wilder. They are
not typical according to Ferussac's figures and can be considered as a geo-
graphical race.

ShkivIv Characters

Ferussac's figures closely represent most of the specimens collected on
Konahuanui. The almost flat spire, only slightly elevated, the high base
below the peripheral keel, forming an inverted truncate cone, and the shape,
formation, and size of the umbilicus are characteristically portrayed. In
addition to these structural features, the color and markings are similar to
some of the specimens before me. However, the bases of most specimens
in the Museum are covered with a buflPy citrine uniform epidermis, with only
faint indications of maculated lines.

In most specimens of B. lamellosa from Konahuanui with more than
5J^ whorls there is a very narrow spiral terrace from a third of a whorl
to a whorl long. The width of this terrace varies in different specimens.
In all it is formed by the undercutting of the columellar wall bordering the
umbilicus, the outer and upper edge being slightly more contracted than the
inner edge. In Ferussac's figure the umbilicus is approximately 28 per cent
of the greatest diameter. In four adult specimens (Bishop Museum Nos.
14265, 16656, 17606) the umbilicus measured from 27.5 to 29 per cent of
the greatest diameter of the shell.

The height of the spire varies slightly, in some specimens being almost
flat, in others only slightly raised. The whorls as a rule are separated by

Pilshry, Cooke, Neat — Snails from Hawaii, Christmas, Samoa 17

a very superficial suture, for they are united on almost the same plane. In
a few older specimens the last whorl descends slightly for its last third,
overhung by the periphery of the penultimate whorl.

None of the specimens examined has seven shorls, the number given by
Deshayes. The greatest number, 6^ whorls, was found in two individuals
(Nos. 17606 and 14265).

All the specimens of B. lamellosa in the Bishop Museum are apparently
much shorter than the one measured by Deshayes. The greatest height of
any in the Bishop Museum is 3.2 mm. Deshayes gives 4 mm. as the height
of his shell, which is slightly more than half the greatest diameter. A
careful measurement of his figure shows the height to be 44 per cent of
the diameter, which coincides almost exactly with the measurements of the
two largest shells in the Museum, given below. The measurements in milli-
meters of four adult shells are as follows:

CATALOG
NUMBER

he:ight

gre:ate:st

NUMBER Ot

diame:te:r

WHORLS

7.1

6.7
6.8
7.0

6
6

14265 3.2

16656 3.0

17606 3.0

17606 3.1

Unfortunately the folds of Ferussac's much enlarged figure are not
drawn with as much detail as could be wished. The relative position and
size of the folds are well represented; but the details of the form of the
individual folds have not been brought out.

According to Ferussac's figure there are two parietal lamellae, one
columellar fold, four basal and one palatal lamina. Of the 97 specimens
of this species from Konahuanui, in Bishop Museum, only one shows an
adventitious lamella. In this specimen the columellar fold is duplicate. In
all the specimens two parietal lamellae are present, the outer much stronger
than the inner, and with its free margin bifid. The smaller inner lamella is
arcuate in cross section, and bent slightly inward. The columellar fold,
situated near the base of the columella, slopes upward.

There are four nearly equally developed basal laminae, the innermost
sloping outward, the three others nearly perpendicular to the base. The
two inner are inserted at about an equal distance from the margin of the
aperture, the third and fourth at somewhat increasing distances internally.
In addition to these four basal laminae, 72 of the 97 specimens have a
very short, low fold placed farther inwards between the fourth lamina and
the peripheral keel. This fifth fold seems to have escaped the attention of
former observers. In all specimens the upper parietal wall is furnished
with two unequal laminae: a large one centrally situated and a much lower

i8

Bernice P. Bishop Museum — Bulletin 47

short one deeply seated and midway between this lamina and the peripheral
keel.

In Ferussac's figure the columellar fold is placed higher than has been
observed in any Oahuan Bndodonta, showing relationship with lamellosa.
The inner lower corner of the aperture as drawn is similar to that in the
Konahuanui specimens.

A comparison of the specimens in the Bishop Museum with Ferussac's
figures makes me feel quite certain that Ferussac's type colony is either
identical with that from which the Museum specimens were obtained or

2 rr\Yr\.

Figure: 3. — Bndodonta lamellosa Ferussac : a, diagrammatic drawing of a nearly
fully expanded alcoholic specimen, g. p. = genital pore, int. = intestine ; b, end of tail,
showing pedal grooves; c, cross section of foot just back of peduncle; d, cross section
of foot about one-half distance between peduncle and tail; e, cross section of mantle
just above collar, showing indentations of lamellae ; /, pallial organs and alimentary
tract, h. = heart, int. = intestine, k. = kidney, m. = muscles between lamellae, p. p. =
pulmonary pore, p. v. = pulmonary vein, u. = ureter ; g-l, cross sections of penis in figure
5, d-g at I, h at II„ i at III, / at IV, ^ at V, / at VI.

Pilshry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 19

that it is situated at no great distance, probably less than a quarter of a
mile, though it may have been at a somewhat lower altitude. There is little
doubt that both belong to the same geographical race.

So]?T Anatomy

The external portion of the animal is uniformly whitish with a slight
yellowish tinge. The pedal grooves are shallow and very narrow but easily
made out. Their position is shown in figures 3, ^, c, d. The sole is narrow
and nearly smooth. The surface above the pedal grooves is very minutely
granulose. The genital pore is placed very far forward and is almost under
the base of the tentacle (fig. 3, a).

The mantle is thin, colorless, and strongly indented where it meets the
apertural folds (fig. 3, e). The lung cavity is long and narrow. The pul-
monary vein is well defined and extends nearly to the collar. The kidney is
long and narrow, and the urethra crosses over from near its base to under
the intestine (fig. 3, /).

The genital organs are remarkably simple (fig. 5). The penis is long-
cylindric, tapering gradually towards the base, where it narrows rapidly just
above the genital pore. Ordinarily it is widest where the vas deferens
enters and does not taper to any extent above; the distal end is rounded.
Three penes had the following measurements :

Specimen B : length 5.3, diameter 0.8 mm. (fig. 5, a, b)

Specimen A : length 6.1, diameter 0.9 mm. (fig. 5, d)

Specimen C: length 4.4, diameter 0.7 mm. (fig. 5, c)

Figure 4. — Bndodonta lamellosa Ferussac: a, free muscles, p. r. = penis retractor,
p. = pharyngeal retractor, 1. t. r. = left tentacular retractor, 1. o. r. = left ocular
retractor, 1. f. b. = left front body retractor, t. = tail retractor, r. f. b. = right front
body retractor, r. t. r. = right tentacular retractor, r. o. r. = right ocular retractor;
h, teeth, C-8, 13-18, geatly magnified.

20

Bernice P. Bishop Museum — Bulletin 47

The penis retractor is attached to the distal end of the penis. The vas
deferens enters the penis at from a third to a quarter the distance from its
distal end. In cross section the lower portion of the penis is a simple tube,
but upwards two well-developed pilasters are gradually formed (fig. 3,^-0 •

Figure S. — Bndodonta lamellosa Ferussac, reproductive organs: a, lower view, her-
maphrodite organs omitted, Specimen B, alb. = albumen gland, sp. = spermatheca,
prst. = prostate, o. d. = oviduct, v. d, = vas deferens, sp. d. = spermatheca duct,
p. == penis, p. r. = penis retractor; b, reproductive organs, upper view, hermaphro-
dite organs omitted, same specimen, lettering as in a; c, upper view, upper female
organs omitted, Specimen C, lettering as in a; d, lower view, upper female organs
omitted. Specimen A., lettering as in a.

Pilsbry, Cooke, Neat — Snails from Hawaii, Christmas, Samoa 21

The oviduct is simple, long, and narrow, the upper third covered towards
the outside with a well-developed, narrow, granular prostate gland; adjoin-
ing the distal end of the oviduct is a rather large albumen gland. The
spermatheca is pyriform and lies just under the albumen gland. It is fur-
nished with a long narrow duct, which is inserted into the uterus just above
and close to the genital pore.

The penis retractor is attached close to the distal end of the free muscles
and just below the attachment of the pharyngeal muscles. The left tentacular
muscle is attached to the free muscle somewhat more distally than the right
tentacular (fig. 4, a).

The radula is long and narrow. The tricuspid centrals are furnished
with a large square basal plate. There is a gradual transition between the
laterals and marginals. Beginning with the third lateral the diminution in
the size of the basal plate increases. The fifth tooth has the mesocone
slightly notched on its inner edge. The notch increases in size to about the
fourteenth or fifteenth tooth, where the mesocone is divided into two almost
equal cusps. Just beyond these the ectocone splits into two minute cusps.
The three or four outermost marginals are ordinarily very minute squarish
plates without cusps (fig. 4, b).

The tooth formulae of five specimens examined are 1-4-16 (two speci-
mens), 1-4-18 (two specimens), and 1-4-19 (one specimen).

The jaws were very thin and delicate and were not isolated.

Conclusion

For the identification of B. lamellosa the figures published by Ferussac
in the Histoire and Deshayes's French description should be relied on. The
Latin description in the Histoire should be ignored, as it is taken directly
from PfeiflFer, whose description is based on an entirely different species,
probably £. marsupialis, Pilsbry's and Vanatta's figures illustrating E.
lamellosa are also drawn from another species, undoubtedly E. fricki or
one of its close relatives.

ENDODONTA MARSUPIAUS

Endodonta marsupialis Pilsbry and Vanatta

Bndodonta marsupialis Pilsbry and Vanatta: Acad. Nat. Sci. Philadelphia, Proc.
1905, p. 784, 1905-06.

Helix lamellosa Pfeiffer: Helix II, Conch.-Cab. of Martini and Chemnitz, p. 197,
pi. 100, figs. 6-9.— Tryon : Man. Conch., vol. 3, P- 67, pi. 12, figs. 32-34, 1887.— Pfeiffer :
Zeitschr. f. Malak., p. 85, 1845; and Mon. Hel., vol. 1, p. 184, 1848, if these two
descriptions are based on the specimen figured in Helix II, Conch.-Cab., pi. 100, figs. 6-9.

The species Endodonta marsupialis was described by Pilsbry and Vanatta
from specimens collected on Oahu. Dr. Pilsbry kindly deposited two of the

22 Bernice P. Bishop Museum — Bulletin 47

paratypes in the Bishop Museum. On the western slope of Mount Tantalus,
directly above the floor of Pauoa Valley on the trail leading to the southern
extremity of the peak, I located a small colony of the species. In two trips
only 6g live and whole dead specimens were found. Two small series that
came to the Museum with the H. Waterhouse collection were probably
obtained in the immediate neighborhood.

vShkll Characters

The species is easily recognized by its proportionately narrow umbilicus
and rather high spire.

Of the 86 specimens from all sources that I have examined, only 2 have
a fifth minute deeply seated outer basal lamina. In this respect the species
differs from E. lamellosa, in which in nearly three-fourths of the specimens
examined this tooth was present. Somewhat less than half the specimens
(35) have only a single upper palatal plica. The small, deeply seated outer
palatal plica, present in every specimen of £. lamellosa examined, is present
in 51 specimens (59 per cent) of B. marsupidis.

A character that this species has in common with Tahitian and Cook
Islands Libera, but in a less marked degree, is the prolongation of the lower
inner margin of the aperture (below the columellar fold) into a distinct
sinus, which extends to and forms the termination of the angle bordering
the umbilicus. This sinus is merely indicated but not developed in speci-
mens of B. lamellosa.

Adult shells of B. marsupialis are slightly larger than those of B. lamel-
losa with approximately the same number of whorls. Measurements in
millimeters are as follows :

CATAI^OG
NUMBER

HEIGHT

DIAMETER

NUMBER OF
WHORLS

16118
16118
16118 dead
16118 dead

3.6
3.8
3.7
3.7

7.6
7.8
8.1
8.2

63/4

6y2
m
6y2

SoET Anatomy

Externally there is no observable difference between the animals of £.
marsupialis and B. lamellosa. In B. marsupialis the reproductive organs,
proportional to the size of the animal, are much larger. The oviduct is
somewhat longer and is distinctly twisted or kinked. The spermatheca ana
albumen gland are decidedly larger. The hermaphrodite duct is a simple
curved tube, widest near its middle and tapering towards both ends. A
single specimen of the hermaphrodite gland dissected was found to consist
of 20 two- to four-branched follicles.

Pilshry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 23

The penis differs considerably in the two species under discussion. That
of H. marsupialis is long and slender and is ordinarily folded two to three
times. The two specimens figured measured approximately 12.0 by 1.0
mm. and 15.0 by 0.8 mm. (See fig. 6, a,b.) A third specimen, carefully pinned
out, measured 14 mm. in length. In Specimen B (the more characteristic
of the two figured specimens) from the point where the vas deferens enters
it, the penis tapers gradually towards its distal end, where the penis retractor
is attached. Below the entrance of the vas deferens the penis has an almost

Figure 6. — Bndodonta marsupialis Pilsbry and Vanatta : a, b, reproductive organs,
alb. = albumen gland, h. == hermaphrodite duct, o. d. = oviduct, p. = penis, prst. =
prostate gland, p. r. = penis retractor, sp. = spermatheca, sp. d. = spermatheca duct,
t. = talon, V. d. = vas deferens; c, marginal teeth, 14-19, greatly magnified.

uniform diameter to a point directly above the genital pore, where it narrows
gradually. The vas deferens enters the penis about 3.5 mm. from its distal
end. In cross section the penes show an internal formation of pilasters
similar to those of E. lamellosa. In form and size the penis of B. mar-
supialis approaches closer to B. fricki than to E. lamellosa. It differs, how-
ever, from E. fricki in that the vas deferens enters the penis at a propor-
tionately greater distance from its distal end.

The tooth formula of the radula of E. marsupialis differs slightly from
that of E, lamellosa. Three radulae have the following formulas: 1-5-18,
1-5-17, and 1-5-20. All three radulae of E. marsupialis examined have five
laterals; v/hile E. lamellosa has four. The centrals and laterals are both
characterized by rather wide basal plates. From the fourteenth to the
seventeenth marginals the cusps are more numerous and much more strongly
developed in E. marsupialis than in E. lamellosa. (See fig. 6, c.)

24 Bernice P. Bishop Museum — Bulletin 47

C0NCI.US10N

The species E. lamellosa and JS. marsupialis differ not only in shell char-
acters but also decidedly in the size and form of their reproductive organs,
especially the male organs. In the arrangement and form of some of their
teeth are also some slight differences.

ENDODONTA FRICKI

Endodonta fricki Pfeiffer

Helix fricki Pfeiffer: Zool. Soc. London, Proc. 1858, pp. 21-22, pi. 40, fig. 3. —
Tryon: Man. Conch., vol. 3, p. 67, pi. 12, figs. 35 and 36, 1887.

Bndodonta lamellosa Pease: Zool. Soc. London, Proc. 1871, p. 474. — Pilsbry: Man.
Conch., vol. 9, pi. 4, figs. 40 and 41, 1894. — Sykes: Fauna Hawaiiensis, vol. 2, part 4,
Mollusca ; p. 287, 1900. — Pilsbry and Vanatta : Acad. Nat. Sci. Philadelphia, Proc, 1905,
p. 783, pi. 43, figs. 7 and 8, 1905-1906.

The name Helix fricki was proposed by Pfeiffer in 1858. That the
illustration accompanying the description was undoubtedly drawn from a
specimen from the Waianae Mountains is the conclusion reached after a
comparison of both the description and figure with the large series of typical
Endodonta from the Waianae Mountains in the collection of the Bishop
Museum. The several hundred specimens in the series were found in about
67 restricted colonies covering the whole length of the range. Included in
the series are a number of geographical races, whose relationships to what
are apparently distinct species are somewhat involved.

She:ivL Characters

Taken together Pfeiffer's description and figure do not exactly fit any
one of the geographical races represented in the collection of the Bishop
Museum.

The post-nuclear whorls of some of the Waianae forms related to fricki
are indistinctly costulate, the others are very faintly and irregularly striate.
None of the numerous specimens that have come under my observation
has the suture margined nor subexcavated as mentioned by Pfeiffer. Many
individuals have the last whorl slightly bent downward, and in these the
carina of the penultimate whorl overhangs the last whorl. In no specimen
that I have examined has the columellar fold been absent. In some indi-
viduals it is placed near the base of the columella, in others its position is
about midway. I agree with Dr. Pilsbry that the absence of this fold is
piathologic.

The figures published by Pilsbry and Vanatta show the characteristics
of one of the many Waianae forms. None of the specimens in Bernice P.
Bishop Museum with a diameter of 8 mm. has as many as seven whorls,

Pilshry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 25

the number recorded by Pf eiffer. With this diameter the most typical races
have only from 5^ to 6j^ whorls. Specimens from Pukaloa and Mokuleia
in the Waianae Mountains come fairly close in most characters to Pfeiffer's
diagnosis, though they vary somewhat in measurements and in the propor-
tional size of the umbilicus. Measurements in miUimeters are as follows:

I.OC AI.it Y H HEIGHT

diame:te:r

WHORLS

UMBILICUS

P^R CDNT

29

Pukaloa (14291)

2.7

7.2

6+

Pukaloa (14291)

2.6

6.8

6

27

Pukaloa

second colony( 16019)

3.1

8.0

6y2

36

Pukaloa

second colony (16019)

3.5

8.7

6V4

40

Mokuleia (17215)

3.4

8.7

6^

28

Mokuleia (17215)

3.8

8.2

6%

25

The two colonies from Pukaloa are not more than a few hundred yards
apart, and yet their shells show considerable differences in the proportional
size of the umbilicus. In the second colony the size of the umbilicus
approaches that given by Pfeiffer.

In 44 specimens (catalog No. 14236) of E. fricki from the same colony
as 14291, 36 (82 per cent )had but four basal laminae, 8 (18 per cent) had
a fifth small outer deeply seated basal fold. In 92 specimens from Mokuleia
(No. 17215) 54 specimens (59 per cent) had four basal laminae and 38
(41 per cent) had the additional basal fold. In the absence of the outer
fifth basal fold most specimens of this species differ from lamellosa, in
which this fold appears in nearly 75 per cent of the specimens. This tooth
is more common in fricki than in marsupialis. Though of the 44 specimens
in No. 14236 no adventitious lamellae were found, 11 specimens in No.
17215 showed abnormal lamella arrangement due to the addition of adven-
titious lamellae. In two specimens from this colony as many as 12 lamellae
are present. The columellar fold is ordinarily situated slightly below the
middle of the columella, in a few specimens almost basal in position. This
character may have led Pfeiffer to state that the columellar fold is absent.
As in lamellosa, there are always two upper palatal laminae, a large one
centrally placed and a short, low one deeply seated, midway between the
central and the peripheral keel. The more prominent upper palatal fold is
characteristically formed in all the geographical races of E. fricki from the
Waianae Mountains. Its free margin is abruptly bent outwards almost at
right angles, so that in cross section it is L shaped. This character does not
pccur in either of the two species from the Koolau Range.

Soi^T Anatomy

Six specimens of one of the geographical races of £. fricki from Lua-
lualei, on the south side of the Waianae Mountains, were dissected. The

26

Bernice P. Bishop Museum — Bulletin 47

outside of the animal is similar to that of lamellosa and marsupialis. Some
rather important differences exist in the reproductive organs. The prostate
gland is proportionately longer and narrower than that of lamellosa. As in
marsupialis, the penis is kinked or twisted. In length it is 9 to lo mm. and
in size is intermediate between that of lamellosa and marsupialis. Unlike
either of the other two species, the vas deferens of fricki enters the penis
close to its distal end. In cross section the interior shows about the same
muscle arrangement as in lamellosa. Another constant character found in
this species is the numerous kinks or folds in the vas deferens in the region
near the genital pore. (See fig. 7, a, h, d.)

c

Figure 7. — Bndodonta fricki Pfeiffer: a, reproductive organs, upper female organs
omitted, o. v. = oviduct, p. = penis, prst. = prostate gland, p. r. = penis retractor,
sp. d. = spermatheca duct, v. d. = vas deferens; b, male reproductive organs, letter-
ing as in a; c, central tooth, greatly magnified; c?, hermaphrodite organs, h. d. = her-
maphrodite duct, h. g. = hermaphrodite gland, t. = talon.

The teeth of B. fricki are much smaller than those of E. lamellosa. In
the two radulae examined the tooth formulae were 1-7-14 and 1-9-12. There
are nearly twice as many laterals as in lamellosa. The basal plate of the
centrals is proportionately much narrower, its width not quite equaling its
length, whereas in lamellosa the basal plate is much broader than long. In
both specimens of £. fricki the mesocone of the centrals extends slightly
below the middle of the plate. The back of the centrals is not as concave

Pilsbry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 27

as in lamellosa (fig. 7, c). The side cusps of the centrals and the ectocones
of the laterals and marginals are very minute, showing merely as points of
light with the same magnification as that used in studying the radulae of
lamellosa.

Conclusion

From the characters given in Pfeiffer's diagnosis and in his figures I
feel certain that his original specimens were from the Waianae Mountains
of Oahu. Specimens agreeing exactly in all characters with his description
and figure are not present in the collection of the Bishop Museum. Speci-
mens from three colonies, however, agree closely enough to be chosen as
representatives of his species. Not only the characters of the shells but
also the anatomical diiTerences lead me to believe that B. fricki should be
considered distinct from B. lamellosa.

BIBLIOGRAPHY

1. Deshayes, G.-P., Hist. Nat. Moll., vol. 1, p. 369.

2. Ferussac, D. Dt, Hist, Nat. Moll., pi. 51 A, fig. 3.

3. Pease, W. H., Catalogue of land shells inhabiting Polynesia: Zool. Soc. London,

Proc. 1871, p. 474.

4. PeeiEEER, L., Zeitschr. f. Malak., p. 85, 1845.

5. PeeiEEEr, L., Helix II : Conch.-Cab. of Martini and Chemnitz, p. 197, pi. 100,

figs. 6-9.

6. Peeieeer, L., Mon. Hel., vol. 1, p. 184, 1848.

7. PeeiEEER, L., Descriptions of eleven new species of land shells from the collection

of H. Cuming, Esq.: Zool. Soc. London, Proc. 1858, pp. 21-22, pi. 40, fig. 3.

8. PiLSBRY^ H. A., Guide to the study of Helices: Man. Conch., vol. 9, pi. 4, figs.

40, 41, 1894.

9. PasBRY, H. A., and Vanatta, E. G., Hawaiian species of Endodonta and Opeas:

Acad. Nat. Sci. Philadelphia, Proc. 1905, p. 783, pi. 43, figs. 7, 8, 1905-1906.

10. SykES, E. R., Fauna Hawaiiensis, vol. 2, part 4, Mollusca, p. 287, 1900.

11. Tryon^ G. W., Helicidae: Man. Conch., vol. 3. p. 67, pi. 12, figs. 35, 36, 1887.

28

Bernice P. Bishop Museum — Bulletin 47

DISTRIBUTION AND ANATOMY OF PUPOIDOPSIS
HAWAIIENSIS

BY C. MONTAGUE COOKE:, JR., AND MARIE C. N^AI,

DISTRIBUTION

The only known species of Pup old op sis, P. hawaiiensis, described by
Pilsbry and Cooke,^ is abundant on all the main Hawaiian islands wherever
eolian calcareous "sandstone" occurs. In addition to the localities reported
in the Manual of Conchology, specimens were taken later at Koloa,
Mahaulepu, and Kealia on the Island of Kauai. All Hawaiian specimens
are fossil. While on the Whippoorwill Expedition in 1924, T. T. Dranga

Figure 8. — Pupoidopsis hawaiiensis Pilsbry and Cooke: a, shell (No. 794^7) with
expanded animal, penis, and tentacles, f. = foot, gen. ap. = genital aperture, i. t. =
inferior tentacle, o. t. = ocular tentacle, p. = penis ; b, pallial organs and alimentary
tract (No. 79490), a. = anus, h. = heart, int. = intestine, k. = kidney, 1. == lung,
liv. = liver, p. v. = pulmonary vein, st. = stomach, u. = ureter, greatly magnified^
c, alimentary tract (No. 79490) showing stomach, greatly magnified.

found living specimens on Christmas Island, more than 1,200 miles distant
from Hawaii. He collected them on bunch grass (Lepturus repens), on
Which they were very abundant, and later in the year S. C. Ball and G. P.
Wilder collected more in several localities on the same island. The shells

•Pilsbry, H. A., Pupillidae: Man. Conch., vol. 26, pp. 106-108,1920-1921.

Pilshry, Cooke, Neal — Snails from Hawaii^ Christinas, Samoa 29

agree with Hawaiian fossils and suggest the probable habitat of this species
in the Hawaiian islands, where with the introduction of cattle and goats
the bunch grass gowing on sand dunes has been destroyed. Thus since the
advent of the white man the abundant fossil material has come into exist-
ence. The most recent Hawaiian shells were found on Molokai under frag-
ments broken from ancient consolidated dunes. The absence of this species
from Kahoolawe and Hawaii is probably due to the absence of such dunes
on these two islands. It is found under similar conditions^ to the nearest
related genus, Pupoides, a cosmopolitan genus.

The distribution of Pupoidopsis hawaiiensis on two land areas widely
separated by ocean is remarkable. Its presence on Christmas Island is
probably due to early Polynesian voyages from the Hawaiian islands, who
do not seem to have carried it farther south, as the species has not been
found on Maiden or Penrhyn islands, where S. C. Ball made a special
search for it. Neither has it been found on Palmyra, Washington, and
Fanning islands, although many specimens of minute species of shells from
these three islands are in the collection of the Bishop Museum.

ANATOMY

Judging by shell characters only, Pilsbry included Pupoidopsis in the

subfamily Pupillinae, a classification that is further proved correct by the

anatomy. For inferior tentacles are present and also an appendix to the
penis.

The expanded animals (fig. 8, a), prepared by drowning in water and
then preserving in alcohol, are light-slate on the sides of the foot and
somewhat lighter on the sole. Indications of pedal grooves are lacking,
and the sole is entire. In expanded animals the genital pore is midway
between the inferior tentacles and the collar and just above the upper margin
of the foot. The breathing pore is located in the collar above the right
tentacle and is remarkably large. The ocular tentacles are somewhat darker
than the foot, and in no specimen were the inferior tentacles expanded,
though their position was clearly indicated (figs. 8, a and 10, a) . The exterior
edge of the mantle is very dark slate. The wall of the lung is very thin
and transparent, in some specimens scarcely pigmented, m others with lines
of minutely pigmented dots accompanying the intestine, and in some bor-
dering the lower margin of the kidney, the pericardium, and the pulmonary
vein. The folds of the kidney, the position of the pulmonary vein, and the
heart are easily seen through the very thin lung wall.

' op. cit., p. 109.

30

Bernice P, Bishop Museum — Bulletin 47

Jaw and Teeth

The jaw (fig. lo, c) is simple and semicircular and does not appear to
have distinct vertical or arcuate striations. It is extremely tough and
resists the action of potash when organic matter is being dissolved from the
radulae. The radula is composed of almost transverse rows of teeth. The
formula is 9-5-1-5-9 (fig. 10, d). In some specimens are six laterals. The

app.

Figure: 9. — Pupoidopsis hawaiiensis Pilsbry and Cooke (No. 79490 ) : a, male and
female genitalia: alb. g. = albumen gland, app. = appendix, h. d. = hermaphrodite
duct, h. g. = hermaphrodite gland, o. d. = oviduct, p. = penis, r. p. = retractor penis,
sp. == spermatheca, sp. d. = spermatheca duct, ut. = uterus, v. d. = vas deferens,
V. s. = vesicula seminalis; b, female genitalia: 1. alb. g. = lobe of albumen gland,
other lettering as in a; c, detail showing hemaphrodite organ, h. d. = hermaphrodite
duct, h. g. = hermaphrodite gland, v. s. = vesicula seminalis.

centrals are minute, very indistinctly tricuspid with a narrowly oblong basal
plate. The laterals are of the ordinary form and are bicuspid, the larger
cusp extending to the base of the quadrate plate. The first marginal is a
transition between the laterals and marginals. In this tooth the ectocone is
ordinarily bifid. In the second, third, and fourth marginals the large cusps
diminish and the outer cusps increase. The third marginal is indifferently
three- or four-cuspid. Most of the outer marginals have four more or less
equal cusps. The teeth have a superficial resemblance to those of Thaanumia
fusculaJ

^Pilsbry, H. A., and Cooke, C. M., Appendix to Amastridae. Tornatellinidae : Man. Conch.,
vol. 23, pi. 21, fig. 11, 1915-1916.

Pilsbry, Cooke, Neat — Snails from Hawaii, Christmas, Samoa 3 1

Kidne:y

The kidney (fig. 8, b) is long and wide, diminishing gradually in width
and ending in a rather short ureter. The several longitudinal folds of the
kidney are very distinct. The ureter is much shorter than in Achatinellidae
or Amastridae, and it terminates a short way back of the collar. A white
line in the same position as the fleshy ridge in Achatinellidae accompanies
the ureter on the side opposite the pulmonary vein, which accompanies the
ureter without visible branching,

Genitaua

The male and female genitalia (fig. 9, a) open into an atrium and
are slightly separated from each other. The right tentacle apparently passes
between the penis on the inner side and the oviduct on the outer side. As
in other Pupillids that have been examined, few animals have male genitalia.
In most specimens dissected only female organs were present. The her-
maphrodite gland consists of a single cluster of follicles radiating from the
end of the hermaphrodite duct in both female and hermaphrodite specimens.
The hermaphrodite duct is pigmented with black, its upper portion simple,
its lower portion somewhat convoluted, which is different from specimens
of Sphyradium described by Hanna.^ The proximal end is swollen and
white and, as stated by Watson^ in regard to a similar organ ia genera
belonging to another subfamily of Pupillidae, probably functions as a vesi-
cula seminalis (fig. 9, c). The albumen gland is very large, in some speci-
mens pigmented inside and outside with black in a network pattern. In
the five female specimens dissected a flap was present (fig. 9, ^) below the
point of entry of the hermaphrodite duct and was absent in the four
hermaphrodite specimens. From the narrow unconvoluted uterus the long
narrow oviduct tapers towards the atrium, where it enlarges rather abruptly.
The spermatheca is narrowly ovate, rather short stalked, about half as long
as the oviduct, into which it apparently empties about one-sixth of the length
of the oviduct back of the atrium. As far as the base of the albumen gland
the female organs are enclosed in a tight, darkly pigmented sheath. A
prostate gland was not discovered. The vas deferens passes down the
oviduct, closely adhering to it, from the vesicula seminalis, with which it
seems to be connected (fig. 9, a). It crosses the atrium and from there
passes in a long loop, about twice the length of the penis, to the point of
entry near the upper end of the penis. In female specimens the vas deferens
seems to end near the lower part of the oviduct. The penis is simple and

■ Hanna, G. D., The American species of Sphradium: U. S. Nat. Mus., Proc, vol. 41, p.
375, 1911.

' Watson, Hugh, Masculine deficiencies in the British Vertigininae : Malacological Soc. I^ondon,
Proc, vol. 15, pp. 270-271, 1923.

32

Bernice P. Bishop Museum — Bulletin 47

cylindrical; it narrows just above its junction with the appendix and is
slightly swollen above. The penis retractor is short and seems to be attached
to the lung cavity just behind the collar. The appendix has a total length
nearly equal to the total length of the penis. It is simple and slightly
contracted near the middle. In a few specimens the penis is extruded
(fig. ID, h), w^hen it appears as a simple, straight or slightly arched sac, so
transparent that vas deferens, retractor muscle, and appendix are plainly
visible through it.

In comparison with genitalia of three genera of Vertigininae studied by
Watson, 1^ those of Pupoidopsis differ considerably. It has a proportionately

Figure lo. — Pupoidopsis hawaiiensis Pilsbry and Cooke: a, ocular and inferior
tentacles (No. 79467), cer. g. = cerebral ganglion, i. t. = inferior tentacles, i. t. n. =
inferior tentacular nerve, i. t. r. = inferior tentacular retractor, o. t. = ocular tentacle,
o. t. n. = ocular tentacular nerve, o. t. r. = ocular tentacular retractor ; 6, male and
female genitalia with penis partly extruded (No. 7949o), lettering' as in fig. 9, a; Cy jaw
(No. 79490) (x 80) ; d, teeth (No. 79490).

larger albumen gland; narrower uterus; smaller hermaphrodite gland with
terminal follicles in rosette formation, not clustered on one side of the end
of the duct as in the genera studied by Watson ; larger and lower vesicula
seminalis; longer vas deferens, except in Vertigo, which has one of about
the same length ; and shorter spermatheca duct with lower attachment. The
oviduct is not swollen. The penis is proportionately about the same length
as that of Vertigo, but diflFers from that genus and from Columella and
Truncatellina in tw^o very evident ways: by having an appendix of sub-
stantial size (Vertigininae have none) and by having the openings of the
male and female genitalia slightly separated in the atrium (they are joined
a short way in Vertigininae).

1* op. cit., pp. 278-279.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 33

SUMMARY

Not only in shell characters but also in anatomy is Pupoidopsis closely
related to Pupoides, a genus in the same subfamily. To genera in the sub-
family Vertigininae Pupoidopsis is less closely related.

As in some Vertigininae, at least, the majority of specimens of Pupoidop-
sis hawaiiensis, well developed otherwise, lack male organs.

Female specimens diflfer from hermaphrodite by lacking penis and
appendix and by having apparently a rudimentary vas deferens and a flap
on the albumen gland.

34

Bernice P. Bishop Museum — Bulletin 47

ANATOMICAL STUDIES OF ACHATINELUDAE

BY marie: C. NE:AI,

COMPARISON OF REPRESENTATIVE SPECIES OF THE FAMILY

In their monograph on the Hawaiian family of land mollusks, Acha-
tinellidae, Pilsbry and Cooke^^ dealt exhaustively with the shells. At two
other times^^ pilsbry examined the animals of thirteen species or varieties,
as follows:

Parts figured

Male and female genitalia

Male and female genitalia

Island
Molokai

Hawaii

Name
Newcombia philippiana (Pfeiffer)

Section Baldwinia
Partulina confusa (Sykes)

Section Partulinella
Partulina redfieldii var. kamaloensis

Pilsbry and Cooke
Partulina dolei (Baldwin)

Section Bulimella
Achatinella abbreviata Reeve
Achatinella viridans Mighels
Achatinella byronii (Wood)
Achatinella lila Pilsbry

Section Achatinellastrum
Achatinella stewartii (Green)
Achatinella vulpina (Ferussac)

Section Achatinella
Achatinella lorata var. nobilis Pfeiffer
Achatinella vittata var. simulans Reeve
Achatinella mustelina Mighels

He compared the different organs, especially the genitalia, and found
them so uniform that even the genera could not be distinguished. But as
few specimens of a limited number of species were examined the question
was not definitely answered : Do genera and species of Achatinellidae differ
anatomically ?

In an effort to answer this, examination was made of 97 specimens,
including from one to four species of all sections of all genera, as shown
in Table 5. The species were compared in a general way as to appearance
and in d etail as to shape of kidney, length of attachment of buccal and

"Pilsbry H. A and Cooke, C. M., Achatinellidae: Man. Conch., vol. 22, 1912-1914.
,« ^ ^^"1' APP^"1^^ *o Amastndae.^ Tornatellinidae: vol. 23, pp. 57-60, pis 12-14 20 101c-
Phit7Jnl-^''^P ?• ^" ^"^ '^? zoological position of Parttilina' ani Achatinella: Acad. Nit. Scl
FhUadelphia, Proc. 1900, pp. 565-567, pi. 17, 1901

Male and female genitalia

Molokai

Male and female genitalia,
kidney, free muscles

Maui

Male genitalia

Male genitalia

Male and female genitalia

Male and female genitalia

Oahu
Oahu
Oahu
Oahu

Male genitalia
Male genitalia

Oahu
Oahu

Male and female genitalia
Male and female genitalia
Male and female genitalia

Oahu
Oahu
Oahu

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 35

tentacular muscles, length of penis, length of appendix, diameter of prostate
gland, and distance of beginning of spermatheca duct from the angle the
oviduct makes with the penis. To all specimens examined the following
statements apply:

1. In sinistral snails the genital pore was on the left, and the left
tentacle passed between the penis and oviduct. The reverse was true in
dextral snails. (No abnormally sinistral or dextral snails were dissected.)

2. Contraction of an animal into itself did not cause different measure-
ments in organs from those of fully expanded animals.

3. The organs of pregnant specimens did not have different measure-
ments from those of well-developed adults.

4. Each juvenile snail, with at least two whorls less than the adults,
had proportionately to the size of the animal a much smaller prostate gland,
much shorter appendix and penis (a condition that led to the investigation
described on pages 44-49).

Figure 11. — a, Partulina redfieldii (Newcomb), pallia! organs and alimentary tract,
a. == anus, fl. r. =» fleshy ridge, int. = intestine, k. = kidney, 1. = lung, per. = peri-
cardium, pul. V. = pulmonary vein, u. = ureter; h, Partulina homeri (Baldwin), pallial
organs and alimentary tract (incompletely stretched) ; c, Partulina virgulata (Mighels),
pallial organs and alimentary tract; d, Partulina mauiensis (Pfeiffer), pallial organs
and alimentary tract.

The shape of the kidney varies slightly in width in all sections of
Achatinella and Partulina, some species of different genera agreeing more
closely than species of the same genus. (See figs. 11, 12.) In Newcombia
the kidney was distinctly proportionately shorter. (See fig. 12, a.)

The lengths of attachment of two free muscles, buccal and tentacular,
varied somewhat but overlapped in different species, depending on amount
of contraction or expansion. In some Partulina these two muscles were

36

Bernice P. Bishop Museum — Bulletin 47

attached for a longer distance above the distal end than in Achatinella, in
most of which they were united closely for a short distance, loosely for a
longer, or loosely all the way. An exception to this was P. mighelsiana,
in which the two muscles were not attached.

Figure 12. — a, Newcomhia canaliculata (Baldwin), pallial organs and alimentary
tract, the kidney appears narrow where bordered by pericardium, which overlies and
partly conceals the actually broader end; h, Achatinella lehuiensis var. meineckei Pilsbry
and Cooke, pallial organs and alimentary tract.

The male genitalia (figs. 13-15, 17) varied in size. With some excep-
tions the penis ranged in length from 6 to 13 mm., in different species over-
lapping, being shortest in Partulina rufa (6 to 8), Achatinella vulpina (6 1/2
to 7 1/2), A. lehuiensis (5 1/3 to 6), and A. stewartii var. producta c. f.

v.s.

Figure 13. — Partulina redfieldii (Newcomb), genitalia, app. = appendix, h. d. =
hermaphrodite duct, o. = oviduct, p. == penis, pr. g. == prostate gland, r. p. == retractor
penis, sp. = spermatheca, sp. d. = sperrnatheca duct, ut. = uterus, v. d. = vas defer-
ens, V. s. = vesicula seminalis.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 37

dunkeri (6 to 6 2/3). The exceptions were three species in the section
Perdicella (3 to 5), Partulina horneri (3 1/3 to 4), Achatinella viridans
(4 to 7), three species of Newcombia (2 1/3 to 4 1/3). The exceptions
have the smallest or narrowest shells, excluding P. horneri, which has one
of the largest and widest, and A. viridans.

The location of the penial retractor muscle was quite uniform, slightly
on one side of the tip of the penis. Exceptions were: two Ehurnella,
slightly more on one side of the tip; Partulina dubia, Newcombia cumingi
(fig. 15, d), N. canaliculata, still more on one side. While the location
seemed fairly uniform in all specimens of one species, in three specimens
of N. pfeifferi the three different locations were found.

The branch of the penial retractor muscle between the penis and the
appendix had different sizes and lengths, in the three Perdicella being small,
short, tight, in some webbed to the penis, in some much contracted. As
the webbed condition was found in juvenile Achatinella viridans (fig. 20),
it is perhaps a juvenile character, significant of a low position of Perdicella
in the scale of evolution of the family, a theory borne out by a juvenile
feature of the adult shell, the thin unexpanded lip. In Partulina horneri
this muscle was also short and tight, a state evidently caused by the short-
ness of the first division of the appendix or the low attachment of the muscle
on this division. For in some animals the attachment was near the top, in
others near the middle, in still others near the base of the division.

The appendix showed much variation in different species and owing
perhaps to amount of contraction or expansion considerable variation in the
same species. It has three distinct parts (fig. 14, b). The first, joining
the penis proximally and supported by a branch of the penial retractor, was
shortest in two-thirds of the specimens examined. Ordinarily it is nearly
as wide as the penis. Distal ly it joins the second part, which in one-third
of the specimens was shortest. Ordinarily this part is cylindrical in shape
and tapers rather abruptly to its point of meeting with the long third
division. This last division is a narrow tube of various lengths, and it
gradually widens to about the same size as the second division, forms a
cylinder, and ends in a rounded tip.

The three divisions, measured and recorded separately, had wide and
overlapping lengths in the different species. The first division had a range
overlapping between 2 and 7 2/3 mm., except in the three Perdicella, Par-
tulina horneri, P. nattii, P. proxima, and the three Newcombia, where it
was shorter, and Partulina virgulata, where it averaged longer. The second
division had a range overlapping between 11/3 mm. and 6 3/4 mm., the
three Perdicella and three Newcombia having the shortest. The third
division had a range between 11 and 32 mm., the three Perdicella, Partulina
horneri, Newcombia pfeifferi, and N, cumingi between 4 and 11 mm.

38

Bernice P. Bishop Museum — Bulletin 47

The prostate gland, ranging in diameter from 2 to 6 1/2 mm., was
smallest in the three Perdicella and three Newcomhia, largest in Partulina
rufa and Achatinella stewartii var. producta c. f . dunkeri. In proportion to
the size of the animal it was about the same.

The spermatheca, which the prostate gland encloses, was oval or ellip-
tical except in Partulina horneri, in which it was cylindrical, long, and nar-
row. It varied in size considerably, even in the same species, for example,
in Achatinella viridans and Partulina virgulata.

FiGURS: 14. — a, Achatinella lehuiensis var. meineckei Pilsbry and Cooke, genitalia;
b, Achatinella lehuiensis var. meineckei Pilsbry and Cooke, part of genitalia, showing
three divisions of appendix; c, Partulina virgulata (Mighels), genitalia, one embryo in
uterus, another about to escape from oviduct; d, Partulina rufa (Newcomb), genitalia.

The distance of the beginning of the spermatheca duct from the genital
aperture and its origin on the side of the oviduct varied even in the same
species, for example, in Achatinella vulpina from 2 1/2 mm. to 5 1/3 mm.,
in Partulina nojttii from o to 1 1/3 mm.

Comparison of my figures with those of Pilsbry listed on page 34 showed
close agreement. The drawings of the genitalia of P, confusa gave an
opportunity to compare them with the species of Baldwinia examined by me,
with the result that they seem more like those of dubia than horneri in all
ways except th^ first division of the appendix, which is short, as in horneri.

Pilshry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 39

Genitalia of Newcomhia philippiana resemble those of N. canaliculata.
Pilsbry's drawings of genitalia of three species in the section Achatinella differ
in length of penis and in first division of the appendix, which is very short
in A. lorata var. nobilis. In lorata var. pulchella this was rather long.
His representatives of the section Biilimella agree with my A. viridans
except in the very narrow third division of the appendix of byronii. His
A. stewartii was like my color form of that species, though in the species
the length of the muscle connecting penis and appendix is not clear.

The divisions into sections and genera based on shell characters'^ are
hardly in agreement with the anatomy of those species that Pilsbry and

Figure 15. — a, PartuHna mauiensis (Pfeiffer), genitalia; b, Partulina horneri
(Baldwin), genitalia; c, Partuli^ta dubia (NewcombX, genitalia; d, Newcomhia cumingi
(Newcomb), genitalia.

Cooke and I have examined. Partulina redfieldii '^ and tessellata, Achati-
nella decipens, vulpina, lorata var. pidchella, and turgida were not distin-
guishable from one another. Parttdina proxima was similar except for a
shorter first division of the appendix and a narrow kidney, which resembled
that of F. rufa. P. rufa also differed in the large size of its prostate gland.
P. virgulata differed in the longer third division of the appendix.

13 Pilsbry, H. A., and Cooke, C. M., Achatinellidae: Man. Conch., vol. 22, p. xiv, 1912-1914.

1* An interesting discovery was made in connection with the nervous system of P. redHeldii,
which is, however, unrelated to the matter under discussion — the presence of a slender subcerebral
commissure (fig. i6), which was found in three other Orthurethra by Hugh Watson, See The
presence of a subcerebral commissure in the Orthurethra: Mai. Soc. lyondon, Proc, vol. 15, pp.
280-283, 1923. As it is absent in some of the higher groups of gastropods, the presence of this
nerve in Achatinellidae suggests close relationship of that family with lower groups of gastropods.

40

Rernice P. Bishop Museum — Bulletin 47

<

u

<J

Z
<

a
<;

On

1X4

O

<

s

o

PQ

5h

JO HdVHS

JO XNSJ^HDVXXV

ao a^NNVH

JO XN3[HH0VXXV
JO HXON^l

JO HXONai

^xmxSrSEJJV JO

SNOISIAia
9 JO HXON^l

•ID 5XVXS0HJ

JO astx^Hvm

•a 'Ha^js JO
XDnaiAo NO
jn ^DNvxsia

SOAHSKSl
JO ON

a^NIHVXH

SN^HIDHJS

JO 'ON

QNVISI

<

1^

«-i Co

^-5

-o o

•r ^

1^

a -en

>. i^

O

1

t 7

CO CM

fO 00

CO CO

oj 213 *^0

en en
C G

oo

CO

c4

a

CTJ

J3

§15

C/iPs

^3 *^S 23 oS

nJTJ

Ui<43

COPh C/3Ph

03 0,

a

o
o

3

2 t

3 I

1"

00

mmm tii m m m m m lam

CO

I

J!

CO

.S

CO

CM CM

00

1

.s

<Mr-<

cm"

.S

.-HO
CO

I

CM

1

CM

U

§

CO

2

t

.S.S.S S

T-iCM T-^CM

.s.s .s.s

CM

COrH

3
O

.s y

S
3

B
o

I

m

s

s

t-4

s

(^

i

s

g

s

Pk

Ph

.2

PL,

CO

M

S

"o

§

I 5

€ a

Ph •?
J I

Ph

oJ

Vh CO

Oj W

CM i^

Mi

Ph

OJ

P-i

Pilsbry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 41

o 0^

a>

<u

1 C

S o^

02 0^ £S h

T-2 (U
I"

cool

S ^ "'
.5 • ^-^

Oh ^ a;
O C rt

r^:^ bo

;3 CO

c

<^ -r!

1 1

to

rf*

10

K
K

1

10

o

C/3 CO

e s

1:^ e

p >,

>. >»

^^ >;i

CO CO

-no

:2;0

kJ CO

.2 O

GM p

O-M t^

r-H to

3 3 2 1

"T T I

CX3 to CT)

Csi r-; ^ CM

to CO

t^ VC) 1-H 1—1

tA r6 06 fO

HiiitiAiii!

3

I

J

3

T

7 t "T T 1

10 to 10 CO to

eg r-l

CMr-l

.5 c

C3

CMt-I

^

!s

^

i§

2

;o

'3

^

^

'o

"o

rt

'0

§

§

s

§

§

as

n!

j5

.2

^

rt g

ci

a*

1

s

*co
'a;

6

II
m CO

.s

1

c

1

g

g

§1

.55 Ss

•*-' 'jn

5

3

n

:3

3

^ rt

•^

<u

•4-»

•4->

<*•»

OS

CO

co-g

fi<

a,

A^

A.

P^

<

3

3

3

X!

J3

J=!

rt

rt

rt

S

3

u'u

u

CA <U

S

to

CO

rt

.§

c

ol t»4

13

1

^ c3

iS

^

<:

:=3 3

13

.H

'a3

J

"ci

rt

^

3
XI

o

3

X

3

X

O

s

.^

^v

'C

J4

3

(J

X!

<u

<u

•S

'a>

js

s

^

.s

Jm*

1

g

13

a

.S

rt j2

rs

CtJ ^

bo

a

u Si

i-i

X

o-fi

3

*-:« u

<:

"JS

^

*~4 Q4

1

13 .

13

3 »2

3

•^S

*^

•s

-s

1) M ^

r^ 00

lis

Die

•2 o\ o.

'f i^^-

wi <^ ^ ic

3 O rl-

bO.rT TfPO

X* tS Tf 'T tv
a-^ ^^ ..

C ^ «- t^

" • • .^

»-. ^ \o
"» ^ L^;;o6
.^ "o 2 -"too

-5 *+^ 3^ -Hvo

iJ a c - -^
(u ctf 5.2 nj*i3

.5 «J w a jj o

1 M^ !^A- d ^

^ !> n '^ §^<

<u 2: 3 n 03 -"t
b XI boo ci^

ctf a> ctf .^.j>
^^ ttfP^ « «

<

< <

42 Bernice P. Bishop Museum — Bulletin 47

In the Manual of Conchology^^ the statement is made

Borcherding tentatively suggests the view that P. rufa is "not a pure species, but
a bastard form between P. tessellata Newc. on the one side, and P. proxima Pse. on
the other." While this seems rather doubtful, and probably could be proved only by
breeding experiments, rufa certainly unites characters of the two species. The embryo
has protractive stripes as in P. proxima. Some very rare color forms closely resemble
proxima in the later stages; but it differs from that species by the dull surface, deeply
sculptured spirally; proxima being glossy, with weak spirals and very slight trace of
the minute, wavy, spiral lineolation of the full Partulinas.

This, of course, deals with the shells only. Anatomically P. rufa and
P. proxima were found to be closer to each other than to any other Par-
tulina.

Two sections of Achatinella, represented by Achatinella viridans, lehuien-
sis var. meineckei, and stewartii var. pro duct a c. f. dunkeri, were not dis-
tinguishable except in the smaller size, velvet-black mantle, and gray penis
and appendix of A. I. meineckei.

The section Eburnella, similar in many ways to Partulina redfieldii,
was not found to be uniform. For P. mighelsiana and P. mighelsiana var

Figure i6. — Partulina redfieldii (Newcomb), nerve collar, b. n. = buccal nerve,
cer. g. = cerebral ganglion, ped. g. = pedal ganglion, sub. com. = subcerebral commis-
sure, t. n. = tentacular nerve. Greatly magnified.

bella though alike otherwise differed distinctly in the length of the first
division of the appendix, that of the species ranging from 6 to 7 mm., of
the variety from 3 to 4 mm. P. nattii differed even more widely. The first
division of its appendix ranged from I/2 to 1 J4 nim., the second division
and penis were shorter, and the kidney was distinctly wider than were corre-
sponding parts of the preceding species and variety.

The section Baldwinia showed even less uniformity than Eburnella.

The three species in the section Perdicella showed fair uniformity, Par-
tulina mauiensis differing from the other two in having a shorter third

»5Pilsbry, H. A., and Cooke, C. M., Achatinellidae : Man. Conch., vol. 22, p. 30, 1912-1914.

Pilshry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 43

division of the appendix and a shorter kidney. In length of penis, of
branch of penial retractor, and of appendix they differed from Achatinella.
This is true also of the three Newcomhia, which differed slightly among
themselves. Perdicella and Newcomhia thus seem to be the only distinct
groups in the family.

Summary

Though for the most part alike anatomically, according to data in
Table 5 and figures 11 to 15, a few distinct differences were found in
Achatinellidae, as follows :

1. All species of Newcomhia had comparatively shorter kidneys than
Achatinella and Partulina.

2. Only in Partulina mighelsiana were the two free muscles, the ten-
tacular and buccal, entirely disconnected from each other. In all other
species the length and closeness of attachment varied, but not uniformly.

3. The penis varied in length, being proportionately shorter in the
three Perdicella, in the three Newcomhia, and also, strangely, in large Par-
tulina horneri than in the other Achatinellidae.

4. The location of the penial retractor muscle deviated little from
slightly to one side of the tip of the penis. The shortened or webbed nature
of the branch of the penial retractor in Perdicella is added proof of the
primitiveness of that section.

5. The length of the appendix with its three divisions bears out in a
general way the relationships designated in one or more other organs.

6. In proportion to the size of the animal the prostate gland was about
the same size.

7. The spermatheca was oval or elliptical in all except Partulina horneri.

8. The place of origin of the spermatheca duct on the oviduct was not
constant, even in one species.

9. Divisions of the family by shell characters and anatomical characters
do not agree. Jud^-ing from the few specimens examined, some species had
wide differences in measurements and in shape of some parts of their
anatomy, greater proportional differences than between some of the different
species. The overlapping of variations shows that anatomically Achatinella
and Partulina cannot be differentiated, except in the section Perdicella and
the genus Newcomhia. The consistent differences of the three Perdicella
from other Achatinellidae suggest that the section^^ should be raised to
generic rank. The decision that Newcomhia should have generic value^^ is
confirmed.

1^ op. cit., p. 15.
^"^ op. cit. p. 1.

44 Bernice P. Bishop Museum — Bulletin 47

DEVELOPMENT OF GENITALIA IN ACHATINELLA VIRIDANS

While looking for differences in anatomy in different species, I noticed
the disproportionate size of genitalia in comparison to the size of juvenile
and adult animals, and this led to an investigation of the development of
genitalia in one species. Descriptions and figures of adult genitalia of sev-
eral species in the family Achatinellidae have already been published. (See
page 34-)

A series of the species of land mollusk, Achatinella viridans, was collected
November 26, 1926, in Waialae Iki, Oahu, Hawaii, by William H. Meinecke
for Bishop Museum for use in anatomical studies. The material, consisting
of sixty specimens ranging in size from newly born infants to gravid adults,
was divided into six groups according to number of whorls, which ranged
from 3 to 6 1/4. In each group two or more animals were examined. (See
figs. 17-22.)

Group A

The first animal examined (figure 17) was a gravid adult in the group
having shells with from 5 3/4 to 6 1/4 whorls. All these snails were fully
adult, as evidenced both by the animals and the shells, the shells having
apertures with thickened lips. As was expected, each part of the genitalia
was found to be fully developed. In a second specimen, the male organs
Were about identical in size and length; the uterus was shorter above the
embryo.

FiGURie 17. — Achatinella viridans Mighels (No. 87423), genitalia, ephebic stage-—
app. = appendix, h. d. = hermaphrodite duct, n. = nerve, o. = oviduct, p. = penis,
pr. g. = prostate gland, r. p. = retractor penis, sp. = spermatheca, sp. d. = sperma-
theca duct, ut. = uterus, v. d. = vas deferens.

Pilshry, Cooke, Neal — Snails from Hawaii, Christmas, Samoa 45

Group B

The second group consisted of adult snails (figure 18) with 5 3/4 to
61/4 whorls, the same number as in the first group. But these had no
embryos, and the lip of the aperture was not thickened. Besides not being
distended, the uterus was shorter, and the spermatheca seemed slightly
smaller. Otherwise the organs were found to be as well developed as in
Group A. As a matter of fact, in both of the two specimens examined the
male organs were larger than in Group A; but whether this was due to
individual variation or to some other cause was not discovered. In some
other species examined, pregnancy did not aflfect the length of the penis.

Figure 18. — Achatinella viridans Mighels (No. 87424), genitalia, ephebic stage,
lettering as in figure 17.

Group C

In the third group (figure 19) a decided difference was evident, and
varying stages of development were found, though the range in the number
of whorls of the shells was only half of one whorl, being between 5 1/4
and 5 3/4, most being 5 1/2. This seemed to be a transitional period — an
extremely short one — when the snails were half way between maturity and
youth, that is, in phases of the neanic stage. ^^. One of the specimens I
examined was quite undeveloped (in the metaneanic stage), the other close
to maturity (in the paraneanic stage).

The specimen in the paraneanic stage (fig. 19, a) had smaller and more
slender genital organs than those in Group B. But evidently they were
nearly ready to function. The uterus was as long but much narrower, and
the spermatheca duct branched off from it slightly higher up. The prostate

1* op. cit., Appendix to Amastridae. Tornatellinidae, vol. 23, p. xi, 1915-1916. This and similar
terms given in the following account were first used by Pilsbry and Cooke.

46

Bernice P. Bishop Museum — Bulletin 47

gland was greatly reduced, and the spermatheca smaller. The penial retractor
muscle was slender. When the dissection was made, the appendix was
lying out straight, following the penial retractor muscle. Near the proximal
end of the free muscles, its tip was bent down, apparently only because the
cavity was not long enough to allow it to lie straight throughout its length.
In all maturer specimens the appendix was found lying curled around in
the region of the penis.

Figure ig.— Ac hatine I la viridans Mighels (No. 87425), «. genitalia, paraneanic
stage, lettering as in figure 17; h, genitalia, metaneanic stage.

The specimen in the metaneanic stage (fig. 19, b) had no genital orifice,
and none was found in still younger animals. It was present in all the
preceding groups. Both male and female organs were much reduced, com-
pared to the paraneanic stage, and were flattened, as they also were in the
younger animals. I found no spermatheca. Possibly it was too minute or
else homogeneous with the soft undifferentiated material of the tiny prostate
gland. The spermatheca duct branched ofif from the uterus distinctly higher
up. The penial retractor muscle was not forked. It continued from the
top of the penis down the side a short way and from there was free to a point
near the base of the appendix.

Group D

Examples of the ananeanic stage (fig. 20) (4 3/4 to 5 whorls) were
found to have quite inconspicuous genitalia. Besides being smaller than

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 47

Group C, another difference was found in the penial retractor muscle. The
base of the appendix was webbed by it closely to the penis, from the tip of
which it extended upwards in a thin thread. The spermatheca duct branched
off even farther up on the uterus. The prostate gland appeared to be a
short wrinkled tube.

Figure 20.— Achatinella viridans Mighels (No. 87426), genitalia, ananeanic stage,
with details, lettering as in figure 17.

Group E

In the next smaller size (figs. 21, 22, h) (4 1/2 whorls), also an
ananeanic stage, the appendix and penis were shorter, and a muscle con-
nection between them could not be distinguished. The prostate gland seemed
similar to that in the preceding stage.

Figure 21. — Achatinella viridans Mighels (No. ^^7427), genitalia, ananeanic stage,
with details, lettering as in figure 17.

48

Bernice P. Bishop Museum — Bitllctin 47

Group F

As well-developed embryos of Achatinella viridans have three whorls
the next stage examined (fig. 22, a) was evidently the youngest phase of the
ananeanic, the shells having 3 to 3 1/4 whorls. The genitalia were minute
and translucent, the uterus being almost transparent and also quite frail
(two specimens breaking in the process of dissection at this point before
a whole one could be obtained). The appendix was absent, and the penis
had an enlarged tip, from which extended its retractor muscle as the finest
of threads. Spermatheca duct and vas deferens had not developed, as the
uterus appeared to be an unaccompanied tube. The part that would have
developed into the prostate gland was not even enlarged, an angle at the
tip of the uterus on one end and the narrower tube of the hermaphrodite duct
on the other showing its position. It was embedded in soft clinging mate-
rial, which it was necessary to remove before the outline of the tube could
be seen. The uterus lay in the same position as at maturity and at its
proximal end the oviduct was attached by a wide base to the wall of the
animal behind the right tentacle, at the place where the genital orifice would
later develop.

p-^S- h.a.

Figure 22. — Achatinella viridans Mighels (No. 87428) : a, genitalia, ananeanic
stage, with details, lettering as in figure 17, ped. r. = pedal retractor, r. t. r. = right
tentacular retractor; b, showing position of genitalia in specimen of same size as that
shown in figure 21, t. = tentacle.

At this stage an important feature was a connection near the base of the
penis. It led from there shortly but indirectly to the right cerebral ganglion,
being a thread-like nerve in a wide filmy sheath, which may also have
served to conduct arteries in this region. The connection was noted in all
the other five groups, but it grew constantly looser and slighter with age
and was quite insignificant in mature animals. In the same region a short
connection was noted with the tip of the pedal retractor muscle, which at
this growth stage stopped near the proximal end of the genitalia.

Pilsbry, Cooke, Ned — Snails from Hawaii, Christmas, Samoa 49

Summary

Conclusions that may be drawn in regard to the genitalia of Achatinella
viridans, a representative of the Achatinellidae, are as follows :

1. They develop to greater proportions during the growth of the animal
than the body of the animal.

2. They are not ready formed with the embryo but develop from simple
to complex structures. In the youngest examined, the organs offered slight
suggestion of their condition at maturity, the hermaphrodite duct, prostate
gland, and uterus seeming to form a continuous undifferentiated tube or
thread.

3. Apparently male and female organs develop equally from the young-
est stage to maturity and are ready to function at the same time. So far as
observed they are present during the entire life of the animal.

AUG28 1944

lilllAlHi

3 9015 02327 1805

Onh

uB^^Rr^-

Rimed by Preservation
1996