province of Quito, presents few variations. The confused heaped blocks, which form the volcanic cones, are of the same mineralogical nature as the rock which is their foundation. These cones and steep mountains were doubtless elevated by elastic fluids, at the points where there was the least resistance. The trachytic rock, which is shattered into innumerable fragments, has, as it was brought to the surface, been elevated by the vapours which were evolved. After the eruption, the shattered rock must have acquired a larger volume, as all the masses could not return to the place whence they came; they, therefore, were heaped over the opening through which the evolution of the gas had taken place.

It is precisely what would happen if we were to dig a well in a hard and compact rock, and then wished to throw in the masses of stone thus obtained; the well would soon be filled, and if we were to continue heaping up the stones in the same direction, a cone would be formed over the mouth of the well, which would be higher the deeper the excavation. In this manner, I think, Cotopaxi, Tunguragua, Chimborazo, &c. have been formed.

The elastic fluids, which, after they had broken up the trachytic crust, opened through it a passage for themselves, might on the surface of the ground unite into more or less considerable cavities, hollows already in existence; and we can thus conceive that the masses of rock, which were at first raised, might afterwards sink and occupy these hollows. Thus a cavity must have been formed on the surface, in place of a cone elevated on the point of eruption. I can thus understand the wonderful sinkings presented by the crater of Rucuipichincha, and also the green lake of the solfatara of Tuqueres, of which I have given, in another memoir, a full description.

I consider, then, the formation of the trachytic cones of the Cordilleras as posterior to the elevation of the mass of the Andes. But these cones are not the newest elevations which occur in these mountains. In the neighbourhood of the highest peaks, namely, those of Cayambé, Antisana, and Chimborazo, small mountains are to be observed, consisting of masses of rock of a newer description, and differing distinctly from the usual trachyte. It is black, porphyritic, and its basis, which includes

crystals of glassy felspar, is coloured by augite; the crystals of felspar are rather rare, and the mass has often the air of a basalt; but I have never seen olivine in it. Sometimes this rock is compact and arranged in prisms; sometimes filled with holes like scoria. One would then be inclined to regard it as lava, if it covered considerable spaces; but it occurs always in frag ments, which rarely have the size of a fist. This rock is evidently of very recent origin. At Chorrera de Pisque, near Ibarra, there is a beautiful colonnade, reposing on alluvium; on the estate of Lysco, this substance, in the condition of fragments, has formed a passage for itself through the trachyte, which has been elevated by its agency. It is there where Humboldt believed he saw a stream of lava (coulée), which had issued from Antisana. In another memoir I have explained the reasons which have induced me to differ from the opinion of my illustrious friend. The extinct volcano of Calpi, placed at the foot of Chimborazo, also consists of this kind of basalt. We visited it on our return from Rio-Bamba.

In the midst of the sand which covers the whole plain of RioBamba, there is a hill of a dark colour, called the Jana-urcu (the black mountain).

At the lower part of this hill trachyte projects through the sand; it is of the same nature as that on which, at some distance, Chimborazo rests. This trachyte seems to have been thoroughly shattered; it is full of clefts and fissures in all directions. The acclivity of the Jana-urcu, towards Calpi, consists of small fragments of the black rock, whose heaping together com pletely reminds one of the stone-eruption of Lysco. It appears, indeed, that this eruption of the Jana-urcu, took place after the deposition of the sand which covers the plain; for in the vicinity of the volcano the ground is covered with black scoriaceous

stones.

Our guides, Indians from Calpi, conducted us to a cleft, where the sound of a subterranean waterfall was distinctly heard ; and, judging from the loudness of the noise, the mass of water must be considerable.

The unproductiveness of the soil, from Latacunga to RioBamba has often surprised me.

I asked myself why the glaciers of the high mountains, which

tower above this district, do not give rise to numerous streams. However, the dryness of this plateau is merely superficial; it is certain that the water of these mountains penetrates the loose earth, and then circulates to a greater or less depth in the interior. The subterranean waterfall of Jana-urcu is a proof of this, and further proofs are afforded by the wells, often very abundant, which are to be seen when descending the deep ravines which furrow the alluvial district of this plateau.

Close to Latacunga, between that town and Cotopaxi, there is a well, which was found a few metres under the surface, during an excavation made in the pumice-conglomerates. It was called Timbo-pollo by the Indians. In reality it is not a well, but a subterranean river; but the water is constantly renewed, and even the direction of the stream can be distinctly observed. The temperature of this stream is 18°.8 cent.; the mean temperature of Latacunga is 15.5 cent.

On the 21st December we returned to Rio-Bamba, where I remained a day, in order to finish the observations I had projected.

On the afternoon of the 23d I left Rio-Bamba, following the route to Guayaquil, where I joined the ship which conveyed me to the coast of Peru.

In sight of Chimborazo I parted from Colonel Hall, whose confidence and friendship I enjoyed during the whole of my residence in the province of Quito. His accurate knowledge of the localities was of the greatest consequence to me, and I found in him an excellent and indefatigable travelling companion. We had both for a long time served the cause of independence. Our taking leave was affecting: there was something which seemed to say we should never meet again; and unhappily this mournful presentiment was too well founded. A few months afterwards my unfortunate friend was murdered in the streets of Quito.-Poggendorf's Annalen der Physick und Chemie, 1835,

No. 2.

Communicated

On the Characters of certain groups of the Class Personatæ. By DAVID DON, Esq., Libr. L. S., &c. by the Author.

A CONSIDERABLE time has elapsed since I first commenced the study of the various orders composing this important class of plants, and my attention having been again more recently directed to the subject, I have thought that the observations which I have made might not prove unacceptable to botanical readers. The very close relationship that subsists between the several families and groups of this class renders the circumscription of them extremely difficult, and, indeed, in one like the present, exhibiting a series of natural affinities almost uninterrupted, we can scarcely hope to make out an absolute definition of any of them. Nevertheless, the division of a large class into smaller groups affords greater facilities to their study, and is advantageous in pointing out their peculiarities of structure, and the degrees of development in the various organs.

In order to render this communication more useful in a practical point of view, I shall proceed to treat of the several families separately, and conclude by giving the technical characters of each.

SCROPHULARINEÆ.

I have had occasion elsewhere to remark of the Rubiacea, that they appeared to constitute a grand central point of union between many families of the monopetalous class; an observation which may be applied with equal justice to the family now under consideration, from the great diversity of form found in it, and from its partaking, both in habit and structure, of those orders to whom it is more immediately allied: for example, the group of Scrophulariea closely approximates them to Verbascinea and Solanea; the Gratiolea connects them with Lentibularia; the Antirrhineæ and Gerardiea with Chelonea and Bignoniacea; the Calceolariea with Gesneriacea; the Euphrasiea with Rhinanthaceae; and their affinity to Verbenacea is clearly established through the Ruddlejea. The Verbascinea approach so closely to the first group as to be only distinguished by their symme trical flowers. The normal Solanea are distinguished by the plaited æstivation of their corolla, mostly regular flowers, and

* Linn. Trans. 17. p. 109.

by their curved embryo; and the Cestrine by their valvate æstivation and regular flowers. As neither of the groups above mentioned conveys an adequate idea of the order, I could have wished, therefore, that the name of Personata had been retained for it instead of that of Scrophularineæ, especially as Scrophularia is to be regarded as an aberrant genus.

RHINANTHACEÆ.

This family, consisting of Rhinanthus, Pedicularis, Melampyrum, and Cymbaria, is intimately allied to the preceding order, but limited as here proposed, it is satisfactorily distinguished by its inverted embryo, with the radicle directed to the opposite extremity of the seed. The Euphrasiece have been included by most botanists in this family, and were they to be retained, the separation between it and the preceding order would be rendered impossible. Cymbaria differs from the rest of the order in the very considerable degree of development of its embryo.

OROBANCHEÆ.

A very natural family, well distinguished from the preceding by their persistent corolla, unilocular ovarium, centrifugal embryo, and peculiar economy, in being for the most part parasitical herbs, destitute of proper leaves, and of the green colour common to other plants. The seeds are attached to the narrow depressed placenta, formed by the united edges of the valves. The embryo is placed in the ventral side of the seed, and at the opposite extremity from the hilum, resembling a minute nearly globular homogeneous mass, and, as far as I have been able to observe, perfectly undivided, and consequently devoid of true cotyledons. In all these respects Obolaria and Tozzia seem to agree, but differ in having proper leaves, and in being terres

trial.

CHELONEÆ.

This group is well entitled, in my opinion, to be retained apart from the Scrophularineæ, on account of the greater degree of development of its embryo, its compressed winged seeds, and of the presence of a rudimentary stamen, all of which characters bring it near to the Pedalinece and Bignoniaceæ, from which it is distinguished by its albuminous seeds.

ARAGOACEÆ.

The single genus constituting this group, is one of the most remarkable in point of structure in the whole class. Its regular symmetrical flowers, the twisted æstivation of its corolla, and. the structure of its seeds, bring it very near to the Polemonia