PACKARD.] MORPHOLOGY OF PHYLLOPODA. 383 
sense-filaments, or what we may call the olfactory filaments, are smaller 
and less numerous than in ZL. gouldit. 
In Estheria mexicana, Plate X XIX, fig. 1, la, 1b, the segments on 
the anterior side are produced into what may be called the olfactory tu- 
bercles, which give a bluntly serrate outline to this side of the antenna 
in contrast to the even opposite or inner side. In fig. 1, the antennal 
nerve (at. n.), where the sense cells and tubercles do not occur, is seen 
passing through the middle of the antenna, and the ends of the fibers 
disappear among the nerve cells, which crowd the olfactory tubercles 
(fig. 1b, ol.). In this species no olfactory filaments were observed to be 
present. In Hstheria compleximanus however, Plate V, figs. 3, 4, they are 
well developed, two or three, and scmetimes more, short filaments arising 
from the tubercles, which are more acute than in H. mexicana (from 
Kansas). The structure of the olfactory papille (ol. pap.) is nearly 
identical with those of Limnetis. The ultimate fibers of the olfactory 
nerves are here plainly seen to enter the mass of nerve cells. 
In Limnadia texana, Plate X XVI, fig. 3, the joints of the antennz 
are more richly charged with nerve cells, which are rather smaller than 
in Kstheria compleximanus, though those of the latter species are larger 
thanin HL. mexicana. The histological structure of the first antenne in the 
present family is quite unlike that of the same appendages in the Apodide 
and Branchipodide ; and reasoning by exclusion, and taking into account 
the fact that this pair of antennze do not project much beyond the edges 
of the valves, and that they are placed very near the jaws and mouth- 
opening, and also bearing in mind the great abundance of the sense- 
cells, we are inclined to believe them to be either olfactory or gustatory 
in function, and that in this family at least the first antennz are mainly 
organs of smell or taste. We have often observed Limnetis gould, swim- 
ming quite rapidly on its back at the surface of the water, apparently 
feeding upon the vegetable matter floating on the surface; during its 
movements it would stop and feed upon some object, as if arrested by 
its smell. The sense lodged in these organs are therefore restricted 
either to the sense of smell or taste, probably the former. 
The finer structure of the antennz of Apus has not been examined, 
and it is probable in an indifferent state. That of the first antenne of 
Branchipodide is quite simple. Plate X XIX, fig. 6, represents the first 
antenna of Branchipus vernalis greatly enlarged. It is simpler in struc- 
ture than in the Kuropean B. stagnalis as figured by Leydig (Sieb. u. 
KOll. Zeits. Wiss. Zool., iii, Pl. VILI, fig. 8), since it lacks the series of 
seven sense filaments ending i in knobs of the Kuropean species, though 
the three terminal sete are “much longer. In our species the antenne 
are seen to be unjointed throughout its whole length. ‘Two nerves, one 
on each side (fig. 6n), and composed of several fibers with here and 
there a ganglion cell, ‘approach each other in the middle of the append- 
age, where they are reinforced by a ganglion cell or two. At n/ two 
nerves are seen passing along the center of the appendage; at g ¢ are 
situated several ganglion cells in the nerves, which finally lose them- 
selves in a terminal mass of small compact ganglion cells, situated at 
the base of the three sete. Leydig only fieures “five cells, where in B. 
vernalis they are smaller and much more numerous. It seems obvious 
that in this family the first antenne only possess the sense of touch. 
On Plate XXXIV, fig. 4, Dr. Gissler has figured the first antenna 
of the larva of Streptocephalus texanus, from Kansas, and it will be seen 
that the histological structure is rather different from that of the adult 
Branchipus. The ganglion cells are more abundant at the base of the 
antenne ; the nerve passes along the center, is reinforced by a few large 
