In figure 2 is represented a typical main branch of a male trunk in full fruit which has just begun to form a second bifurcation. The other main branch, bearing a single large male cone, is not shown. In the large main branch as figured two cones are borne by the left and three by the right of the newly forming branches.

So far as can be seen on the outer surface, these Zamia trunks, as first dug up, are rather smooth, and do not appear at first sight to possess the outer armor of spirally arranged leaf bases so characteristic of other cycadean forms. Just below

a. Upper third of long and slender young trunk with leaves cut away, and the surface lightly cleared of scaly and hairy material to show the slowly disappearing remnants of old leaf bases. Further down, that is on the lower two-thirds of the trunk, the leaf bases are indistinct.

All of the trunk as well as all of that portion of the petioles shown at the summit was subterranean.

b. Transverse section of the same trunk cut further down at thickest point, showing the central pith or medulla, the xylem or wood zone, and the cortical parenchyma irregularly traversed by vascular bundles. There is no armor, this being replaced by a very thin corky layer.

the petioles of the crown of leaves there is a thin and rough covering of scattering scale-like bodies and fine hairy material. Below this the trunks appear to be irregularly ridged horizontally, these ridges soon disappearing, so that the lower half or two-thirds is quite smooth in most places, and the armor of old leaf bases absent.

But that these trunks do not differ greatly from other forms is shown in figure 3. The scaly material below the crown of leaves, and, lower down, the horizontal ridges just mentioned, are the remnants of a true ramentum which is constantly disappearing below as the age and size of the trunk increase. And when this ramental material is lightly scraped off above, preferably from a vigorous young trunk, the somewhat carrot or turnip-like appearance is lost, and the spiral order of the remnants of the leaf bases of former years is clearly to be seen. The trunks of the Floridean cycads are hence, notwithstanding their subterranean habit and unusual appearance, typical in every respect. The slow elimination of the old leaf bases is doubtless due in their case to growth in a protected underground position, and may hence be a trunk habit secondarily acquired. The subordinate position of these, comparatively speaking, dwarf plants, as underbrush in large forests of pines, and in the denser "hammocks" in the case of Z. pumila, has, together with differential climatic change, probably, therefore, resulted in the course of time in some diminution in size, an underground habit, and nearly complete removal of the armor.

4

Unquestionably the most interesting single point observed in the cycads sent by Mr. Dickinson was the presence on one of the cones of a pinnule of normal form and structure which had evidently grown out from beneath the outer hexagonal tip of one of the upper abortive sporophylls. Figure 4 is a faithful drawing of this monstrous cone, so far as I know the second thus far observed to bear leaf-like growths in the Cycades. When unpacked the pinnule it bore was still somewhat green. As the parent plant was in good condition and at once planted, it was thought that its examination could be someZamia floridana what deferred. But unfortunately when the cone was examined a week later, the pinnule, which was at first strongly attached, had loosened, so that its insertion was no longer distinct. The thin sections I made of the sporophyll to which it had

DC. × 12. Miami, Florida, Nov. 15, Monstrous ovuliferous cone,

with sporophyll at (a) bearing a pinnule of the normal structure and form seen in this species.

* An anomalous sporophyll of Zamia Leiboldii Miq., bearing three sporangia, has been figured by Mr. Worsdell, Vascular Structure of the Sporophylls of the Cycadaceæ, pl. xviii, fig. 25 (a), Ann. of Bot., vol. xii, No. xlvi, June, 1898.

plainly belonged showed a strengthening of the vascular bundles next the border in contact with the base of the pinnule. But this part had suffered from wilting, so that I could not get as good evidence of attachment as I could have wished. For the sake of exactness I must explain therefore, that while my figure 5 without doubt represents faithfully the condition of several weeks earlier,

it is not the condition found when the cone was examined.

This pinnule did not therefore grow in the exact position of the ovule. in a normal and fertile

5

FIGURE 5.-Zamia floridana DC.

× 34.

sporophyll, but proxi- Sporophyll (a) and pinnule shown in figure 4. mally on the same bor

der. The two outer angles of the sporophyllar rachis may then be considered as being at once spore- or pinnule-bearing. The evidence that the entire structure is a modified pinnate leaf, just as in the case of the carpophylls of Cycas, is overwhelming.

The first example of such monstrous cycad cones recorded, so far as I am aware, is that of an Encephalartos villosus Lem. described and figured by Sir W. T. Thiselton-Dyer in the Annals of Botany (vol. xv, No. lix, Sept., 1901, p. 549). In this case there is a very extended reversion. The barren sporophylls of the summit of the cone become more and more frond-like, until one of them rises as a distinct once pinnate frond. Although truncated, and reduced in size, this frond presents all the essential characters seen in the frond of E. villosus. The twelve pinnules it bears are expanded and notched in a normal manner. "The other modified carpels present,' says Thiselton-Dyer, "are however so generalized that without the help of the more fully developed leaf their equivalence would be scarcely intelligible. This much is clear: the solid expanded peltate carpophyll is nothing more than a transformed foliage leaf and capable of being replaced by it." The italics are mine.

To

The significance of such structures is unmistakable. speak of these growths as "monstrous cones" is almost misleading. They are simply reversions exhibiting evolutionary stages which may at any time be found in fossilized forms of the ancestral line. When an Encephalartos reverts we find the growth preserving unmistakable characters of its genus and species; and likewise in the quite different form of reversion just described in Zamia the phenomenon falls within the same category. The main specific characters are preserved.

Having already discovered the pollen-bearing fronds of Bennettites (this Journal, March, 1899, and June, 1901), the description by Thiselton-Dyer interested me very greatly. I at once thought it probable that such cones might frequently occur, and that in the case of male cones there must be great likelihood of the reverted fronds eventually being found fertile. A few weeks later the Zamia cone above described was obtained. It is to be hoped that all such cases of reversion may be recorded. They promise to relieve us of the necessity for much speculation, and to be fully as important in our own morphological conceptions as have been the cones of other gymnosperms in the hands of Ĉelakovsky', or probably much more so, because these forms yield us a knowledge of very primitive conditions. We are already in some measure guided when we attempt to form an hypothesis of the manner in which prothallial elimination must have proceeded in some marattiaceous or older fern line ancestral to the gymnosperms. At least the series from ordinary cycadean foliage leaves through the carpophylls of Cycas to the less leaf-like sporophyll of the cone (simply a seed-bearing branch) of Dioon, and the ordinary much altered sporophyll of the other cycadean genera, must be regarded as a connected one. And as I have elsewhere pointed out, the staminate fronds of the Bennettiteæ afford concerning the coördinate changes which took place in microsporophylls an explanatory analogy of the most striking character. Progressive prothallial elimination, with correlated spore differentiation and alteration of the frond-like sporophytes of primitive ferns of the marattiaceous or an ancient allied group, were the basal factors in the evolution of the Cycadofilicinean and Cordaitean alliance. This subject of fundamental importance I shall treat more definitely elsewhere.

The question of the homology or equivalency of the Cycadean ovule to pinnules comes up in this connection. But I certainly think that the testimony of other unusual cones which we are almost sure to find in the course of time will be much more satisfactory than any insufficiently founded speculation, and shall therefore content myself with pointing out the opinion of Thiselton-Dyer (loc. cit.) thus conservatively expressed: “**** an ovule is a sporangial structure, and it is not easy to see anything in a pinna which is in any way comparable to it. Morphological conceptions must not enslave us, and I see no reason why sporangial structures, like buds, may not appear anywhere."

ART. XXVIII.-On Crystals of Crocoite from Tasmania; by R. G. VAN NAME.

A RECENT addition to the Brush Mineral Collection of the Sheffield Scientific School consists of a series of specimens of crocoite from near Dundas, Tasmania, purchased from the Foote Mineral Company of Philadelphia. These specimens, which are interesting on account of the diversity of development of the crystals and the unusual habit which many of them show, include both separate crystals and groups attached to the gangue. The latter is a cellular limonite more or less coated and intermixed with a black oxide of manganese (wad), but showing no trace of lead or chromium minerals other than the crocoite. In the cavities of this material, which has evidently come from a zone of oxidation, the crocoite crystals occur in irregular groups or loosely adherent masses.

Two distinct types of development are shown by the crystals. The habit exhibited by the detached crystals, and by a number of smaller individuals still in position upon the gangue, is quite different from that illustrated by Palache* for crocoite from this locality, and is not mentioned by Daubert in his detailed description of crocoite from Brazil, Siberia and the Philippines. Typical specimens of this habit are characterized by a remarkable elongation of the prismatic faces, the unit prism being the predominant form, and giving the crystals a nearly square cross-section, since mm'=93° 41'. Four of the crystals of this type are from 40 to 64mm in length, with a diameter measured across one of the prism faces of from 1 to 2mm, and a number of others, including several fragmentary crystals, have a relative length only slightly less. The faces of the unit prism are lightly striated longitudinally, but are in other respects smooth and even with sharp and well-defined intersections, the whole development of the prism being generally surprisingly regular. In all cases, as far as was observed, the crystals are of uniform diameter throughout and show no tendency to taper. The terminal faces upon the crystals of this type usually show a high polish and give excellent reflections. They are chiefly domes, but the base c (001) and the pyramid (111) are often present though rarely prominent. No doubly terminated crystals of this type were observed.

With one exception all the faces observed were known forms, of which the orthodome k (101) and the clinodome z (011) are the most persistent. The new form is a clinodome j (032), which was found on but one crystal.

*This Journal, i, 389, 1896.

† Ber. Akad. Wien., xlii, 19, 1860.