of the matrix (hypoderm) becomes penetrated by air-canals which independently open at the surface.
It is time that the air-canals have walls of chitine secreted by the hypoderm but these canals do
not open outwards in pores but join one another inwards to form a tree-like cavity which communicates
with the outside of the gill through a single opening.
I have carefully examined the dorsal chitinous wall of the gill both as seen from without
and in sections, but fail to find pore-canals. The author gives a figure of the structures he thus
designates and I find that his error arose through mistaking very small chitinous hairs for canals
lying in the chitinous wall.
The correllative error in L e y d ig ’s account, namely, that then is no opening at the margin
of the gill, is made evident by sections (figures 4 and 7). It is true that there is no opening
near the articulation of the gill with the basal joint but at a point on the opposite margin of the
gill. That L e y d i g was deceived in regard to this becomes understood by the fact that in the
species he studied the marginal indentation, indicating the position of the opening, is small and
inconspicuous.
Leydig appars to have observed the net-work of air-holding furrows, since he refers to a
portion of the pneumatic part as presenting the appearances of a close net-work; but he was evidently
under the impression that the net-work was inside the gill and formed a portion of the general
pneumatic part.
The In n e r Grills. I find that the figure of a section of the inner gill given by L e y d ig
is applicable to my own preparations excepting as regards the nuclei.. In his figure nucleoli are
represented which I do not find to be present. Further he represents the nuclei as evenly granular
throughout while I find the chromatin to be situated mainly on the side lying toward the cavity
of the gill.
In his description of the inner gill he mentions one structural feature which I am not
able to identify in my preparations. He says:
“Es ist dasselbe nach aussen umgrenzt von einer zarten, doch von Porencanälen durchsetzten,
daher senkrecht streifigen Cuticula, oben und unten.
Sodann erscheint die Zellsubstanz von deutlich streifigem Wesen und starke Vergrösserung
bei gehöriger Aufmerksamkeit lässt wahrnehmen, dass die Streifen von Längscanälen oder Lücken
herrühren, welche das Protoplasma durchsetzten. Die Zellen von der Fläche angesehen zeigen sich
wie von feinsten, dicht beisammenstehenden Löchelchen durchbohrt.“
Contrary to the above I find that the chitine is not penetrated by pores. I find that the
protoplasm, of the hypoderm is striated, as stated by L e y d i g , but I do not find that actual canals
or lacunae are associated with the striae.
Armadillidiwm vulgare Z o d d a c h .
I have examined the first two pairs of outer gills in this species, both as seen from without
and in sections, and I find that in all essential features they are identical with the corresponding
parts in the genera already described. The respiratory tree and grooves present the
same characters in both. The only differences are such as relate merely to comparative form
and size.
Armadillidiwm granulatum B r a n d t .
I have compared alcoholic specimens of this species, collected by Dr. zur Strassen in
Dalmatia, with the preceding species in respect to the external form and structure of the first
two pairs of outer gills. I find that they correspond in all respects, differing only as regards comparative
size and form.
Ligidium Hypnorum B u d d e - L u n d .
In this species the outer gills have no special modifications of structure in adaptation to
the respiration of air. In general they resemble in structure the last three pairs of outer gills in
PorcelUo scdber.
E x t e r n a l A p p e a r a n c e and S tru ctu r e. Fig. 9 shows the appearance of the right
outer gill of the first pair, as viewed on the inner or dorsal surface. It is seen to be of a general
quadrilateral shape and is broader than long. The attachment with the basal joint is at the
outer anterior corner of the gill. The face of the gill is dotted and the dots are so arranged as
to leave open spaces. These are paths for the flow of the blood through the gill. The incurrent
path or channel -fe beginning at the union of the gill with the basal joint — extends laterally
across the gill and gives off branches toward the middle region of the gill. Here the branches
become lost in the small irregular spaces lying among the dots. The excurrent path or channel
lies along the posterior and outer margin of the gill. It has no branches but is in free communication
with the spaces in which the branches of the incurrent channel terminate.
Examining the gill under higher magnification it is seen that the dots are nuclei, lying
immediately within the chitine and belonging (as further shown by the study of sections) to the
hypoderm. By focussing at different levels it may be seen that associated with many of the nuclei
are strands of tissue which run from the upper to the lower layer of hypoderm. These form
supporting columns or pillars, such as have already been described for Porceilio scdber. It is due
to the grouping of these pillars that spaces and channels for the passage of the blood through the
gill, as above described, are formed. Along the channels the pillars tend to be arranged in rows,
forming a kind of broken wall to the channel. Elsewhere they are scattered and stand either singly
or in irregular groups.
I n t e r n a l s tru c tu r e . Fig. 10 represents a cross-section of the outer gill, taken in the
direction of the line 55', Fig. 9. It is seen that the gill is in principle a simple sac containing
blood. The outer layer of the wall, composed of chitine, is much thicker on the ventral than on
the dorsal side of the gill. The hypoderm everywhere lines the chitine and in general varies with
it in thickness. It is very thin on the dorsal side and its nuclei for the most part lie in the
extensions of the hypoderm forming the pillars. The blood-cavity is everywhere bounded by a very
thin wall in which at distant intervals elongated nuclei occur. From its relations I infer that this
layer is mesodermic in origin.1
F u n c tio n of th e o u te r G ills . The structure of the gill as above described appeal's
See Foot-note, p. 14.