(a) the form of both adult and embryo is of a type more adapted for floating than crawling;

(b) the large pygidium is shown by ontogeny to be primitive, and the elongate worm-like form secondary;

(c) the history of the trilobites shows a considerable increase in the average number of segments in successive periods from the Cambrian to the Permian;

(d) the simplest trilobites are nearly or quite blind.

PART IV.

DESCRIPTION OF THE APPENDAGES OF INDIVIDUAL SPECIMENS.

Triarthrus becki Green.

In order to make easily available the evidence on which the present knowledge of the appendages of Triarthrus and _Cryptolithus_ rests, it has seemed wise to publish brief descriptions and photographic figures of some of the better specimens preserved in the Yale University Museum. These specimens are pyritic replacements, and while they do not as yet show any signs of decomposition, it should be realized that it is only a matter of time when either they will be self-destroyed through oxidation, or else embedded for safe keeping in such a fashion that they will not be readily available for study. It is therefore essential to keep a photographic record of the more important individuals.

Specimen No. 220 (pl. 3, fig. 2).

Ill.u.s.trated: Amer. Geol., vol. 15, 1895, pl. 4 (drawing); Amer. Jour. Sci., vol. 13, 1902, pl. 3 (photograph).

This is one of the largest specimens showing appendages, and is developed from the ventral side. It shows some appendages on all parts of the body, but its special features are the exhibition of the shafts on the proximal ends of the antennules, the rather well preserved appendages of the cephalon and anterior part of the thorax, and the preservation of the a.n.a.l opening. In the drawing in the American Geologist, the right and left sides are reversed as in a mirror, a point which should be borne in mind when comparing that figure with a photograph or description.

The shaft of the left antennule is best preserved and is short, cylindrical, somewhat enlarged and ball-shaped at the proximal end. It is 1.5 mm. long. The posterior part of the hypostoma is present, but crushed, and the metastoma is not visible, the pieces so indicated in Beecher's figure being the rim of the hypostoma. Back of the hypostoma may be seen four (not three as in Beecher's figure) pairs of gnathites, the first three pairs broad and greatly overlapping, the fourth pair more slender, but poorly preserved. The inner edges of the gnathites on the right side are distinctly nodulose, and roughened for mastication.

The outer ends of one endopodite and three exopodites project beyond the margin on the right side. The dactylopodite of the endopodite is especially well preserved. It is cylindrical, the end rounded but not enlarged or pointed, and bears three small sharp spines, all in a horizontal plane, one anterior, one central, and one posterior. The outer ends of the exopodites show about ten segments each (in 2.5 mm.) beyond the margin of the test, and from three to five setae attached to the posterior side of each segment. These hairs are attached in a groove, well shown in this specimen. On the anterior margin of the exopodite there is a minute spine at each joint.

_Measurements:_ Length, 38 mm.; width at back of cephalon, 19 mm.

Specimen No. 210 (pl. 2, fig. 3).

Ill.u.s.trated: Amer. Jour. Sci., vol. 46, 1893, p. 469, fig. 1 (head and right side); Amer. Geol., vol. 13, 1894, pl. 3, fig. 7 (same figure as the last); Amer. Jour. Sci., vol. 13, 1902, pl. 2, fig. 1 (photograph).

This individual supplied the main basis for Professor Beecher's first figure showing the appendages of the thorax, the head and appendages of the right side having been taken from it, and the appendages of the left side from No. 206. Such of the endopodites as are well preserved show from three to four segments projecting beyond the test, and the dactylopodites have one or two terminal spines. The antennules are unusually well preserved and have about forty segments each in front of the cephalon, or an average of five to one millimeter.

Specimens 209 and 210 are on a slab about 7 5.5 inches, and with them are twelve other more or less well preserved individuals, all but one of which are smaller than these. Two of the fourteen are ventral side up on the slab, which means dorsal side up in the rock. Nine are oriented in one direction, two at exactly right angles to this, and three at an angle of 45 with the others. If the majority of the specimens are considered to be headed northward, then seven are so oriented, two northeast, one east, two south, one southwest, and one west.

Nine of the specimens show antennules. Five of these are specimens headed north, and in all of them the antennules are in or very near the normal position. The antennules of two, one headed east and the other west, are imperfectly preserved, but the parts remaining diverge much more than do the antennules of those in the normal position. The individual headed southwest has one antennule broken off, while the other is curved back so that its tip is directed northward. Another one, headed south, has the antennules in the normal position. These observations indicate that the specimens were oriented by currents of water, rather than in life att.i.tudes, and that the distal portions of the antennules were relatively flexible.

_Measurements:_ The specimen (No. 210) is 20 mm. long, 9.5 mm. wide at the back of the cephalon, and the antennules project 8 mm. in front of the head. The smallest specimen on the slab is 6.5 mm. long. A specimen 7.5 mm. long has antennules which project 2.5 mm. in front of the cephalon.

Specimen No. 201 (pl. 2, fig. 1; pl. 3, fig. 4).

Ill.u.s.trated: Amer. Jour. Sci., vol. 46, 1893, p. 469, figs. 2, 3; Amer. Geol., vol. 13, 1894, pl. 3, figs. 8, 9.

An entire specimen 17 mm. long, exposed from the dorsal side. It shows only traces of the appendages of the head, but displays well those of the anterior part of the thorax, and a number of appendages emerge from under the abdominal shield. This specimen is of particular interest as it is the subject of the first of Professor Beecher's papers on appendages of trilobites. On the right side the pleura have been removed, so as to expose the appendages of the second, third, and fourth segments from above. The first two of the appendages on the right are best preserved, and these are the ones figured. They belong to the second and third segments. The endopodites of each are ahead of the exopodites, and the proximal portion of each exopodite overlies portions of the first two segments (second and third) of the corresponding endopodite. The c.o.xopodites are not visible, but very nearly the full length of the first segment of the endopodite (the basipodite) is exposed. The first two visible segments (the first and second) extend just to the margin of the pleural lobe, while the other four extend beyond the dorsal cover. The segments decrease in length outward, but not regularly, the meropodite being generally longer than the ischiopodite or the carpopodite. The terminal segment (dactylopodite) is short and bears short sharp hair-like spines which articulate in sockets at the distal end. On this specimen the anterior limb on the right side shows one terminal spine, the second endopodite on that side has two, and two of the endopodites on the left-hand side preserve two each. The segments of the limbs are nearly cylindrical, but the ischiopodites and meropodites of several of the endopodites show rather deep longitudinal grooves which appear to be rather the result of the shrinkage of the thin test than natural conformations.

The endopodites on the left-hand side have a number of short, sharp, movable, hair-like spines, and cup-shaped depressions which are the points of insertion of others. On the distal end of the carpopodite of the first thoracic segment there seems to have been a spine, whose place is now shown by a pit. This same endopodite shows, rather indistinctly, three pits in the groove of the carpopodite, and the propodite has two. On the endopodite of the second appendage on this side, both the carpopodite and propodite possess a fine hair-like articulated spine at the distal end, that of the propodite arising on the dorsal and that of the carpopodite on the posterior side. On the dorsal side of the carpopodite there are three pits for the articulation of spines, and on the propodite, one.

The exopodites belonging to the thoracic segments are of equal length with the endopodites, and while the proximal portion of each is stouter than that of the corresponding endopodite, the exopodites taper to a hair-like termination, while the endopodites remain fairly stout to the distal segment. Most of the setae of the exopodites have been removed, so that each remains as a curving, many-segmented organ, transversely striated, with a continuous groove along the posterior side. The setae appear to be set in this groove, one for each of the transverse ridges on the shaft.

A good deal of the test has been cut away on the left-hand side from the thorax and pygidium, and the appendages exposed from above. Enough of the dorsal sh.e.l.l has been cut away so that the a.n.a.l opening is exposed, and directly behind the pygidium, on the median line, is a bilaterally symmetrical plate with serrated edges which appears to be the appendage of the a.n.a.l segment. (See pl. 3, fig. 4.)

_Measurements:_ The specimen is 17 mm. long, and 8 mm. in greatest width (at the back of the cephalon). From the median tubercle to the outer edge of the pleuron of the second thoracic segment the distance is 3.7 mm. From the point of articulation to the distal end of the spines on the dactylopodite of the second endopodite on the right-hand side is 4.3 mm. The basipodite of this appendage is 1.5 mm. long, the ischiopodite 1 mm. long, the meropodite 1.2 mm. long, the carpopodite 0.5 mm. long, the propodite 0.35 long, and the dactylopodite 0.15 mm.

long. On the left-hand side the endopodite of the first segment projects 3 mm. beyond the pleuron, the second, 3.2 mm. At the back the appendages extend a maximum distance of 2.5 mm. behind the pygidium.

The median spinose process of the a.n.a.l segment extends 0.75 mm. behind the pygidium, and is 1.6 mm. in greatest width.

Specimen No. 204 (pl. 3, fig. 1; pl. 4, fig. 6; text fig. 42).

Ill.u.s.trated: Amer. Jour. Sci., vol. 13, 1902, pl. 2, figs. 4, 5 (reproduced from photographs).

This specimen, which is developed from the dorsal surface, shows especially well nine appendages of the left side. The first represent the last segment of the cephalon; the remainder belong to the thorax.

As is usual, the exopodites of these appendages overlie and curve behind the endopodites. All the exopodites have lost their setae and the segments of the endopodites are flattened by crushing. The endopodites, while retaining only one or two of the movable spines, have the cup-like bases of from two to four on each of the visible segments, namely, the meropodite, carpopodite, propodite, and, in one case, the dactylopodite. These appendages, although really marvellous in preservation, are of such small size and react so badly to light that their study is very difficult, and Professor Beecher, who had observed hundreds of specimens through all stages of the laborious process of cleaning the matrix from them, undoubtedly was much better equipped to interpret them than any other person.

The drawing is made on the a.s.sumption that the appendages are displaced and all moved uniformly outward so that the distal ends of the c.o.xopodites emerge from under the pleural lobe, whereas these ends would normally be under the dorsal furrow, and the distal end of the ischiopodite should reach the margin of the pleural lobe. While it seems very remarkable that it should happen, that all the appendages should be so moved that they would lie symmetrically a few millimeters from their normal position, nevertheless it is found on measuring that they bear the same proportion to the length and width that the appendages of other specimens do, thus indicating that Professor Beecher's interpretation of them was correct. I am unable, however, to see the c.o.xopodites which he has drawn as articulating with the two branches of the limb.

[Ill.u.s.tration: Fig. 42.--_Triarthrus becki_ Green. Appendages of specimen 204. Inked in by Miss Wood from the original tracing. 10.]

This individual shows, better than any other, the connection of the exopodite with the endopodite. Even though the c.o.xopodites are gone, the two branches of each appendage remain together, showing that the basipodite as well as the c.o.xopodite is involved in the articulation with the exopodite. Just what the connection is can not be observed, but there seems to be a firm union between the upper surface of the basipodite and the lower side of the proximal end of the exopodite, as indicated diagrammatically in text figure 33.

_Measurements:_ The specimen is 20 mm. long and 9 mm. wide at the back of the cephalon. From the tubercle on the middle of the first segment of the thorax to the tip of the corresponding appendage the distance is 8 mm. The entire length of the exopodite of the first thoracic segment is 4.6 mm. The exopodite of the appendage belonging to the seventh segment is only 3.5 mm. long. The pleural lobe is 2.5 mm. wide at the front of the thorax.

Specimen No. 205 (pl. 2, fig. 4).

Ill.u.s.trated: Amer. Jour. Sci., vol. 13, 1902, pl. 5, figs. 2, 3 (photographs).

This is a small imperfect specimen, developed from the ventral side.

It retains the best preserved metastoma in the collection, but was used by Professor Beecher especially to ill.u.s.trate the convergent ridges on the inside of the ventral membrane in the axial region of the thorax. These ridges are very low, and on each segment of the thorax there is a central one, outside of which is a pair which are convergent forward, making angles of 35 to 45 with the axis.

The metastoma is shaped much like the hypostoma of an _Illaenus_. It is convex, nearly semicircular, with the straight side forward, and there is a continuous raised border around the curved sides and back. This border is separated from the central convex body by a deep linear depression.

The hypostoma is also rather well preserved and has a narrow, slightly elevated border at the sides and back.

_Measurements:_ The incomplete specimen, from which only a very small portion of the length is missing, is 9 mm. long. The metastoma is 0.45 mm. long and 0.58 mm. wide.

Specimen No. 214 (pl. 1, fig. 2; pl. 3, fig. 6).

This is a large specimen, developed from the ventral side. It shows the antennules and some other appendages of the head, but derives its special interest from the excellent preservation of a few of the exopodites, which are turned back parallel to the axis of the body and lie within the axial lobe.