form a section. Castraea is surely a Phoradendron. Lepidoceras of Dr. Hooker ought not to be merged in Tupeia, notwithstanding the difficulty of finding technical characters whereby to separate it. The habit of the plant is very distinct. It is much nearer to Eremolepis of Grisebach than to Tupeia. The albumen in the seeds of Lepidoceras is very much reduced, while in Tupeia it is copious. The stigma of the latter genus is also capitate. Eubrachion, Hk. f., is a very distinct-looking plant, but we are not in a position, from its extreme rarity, and the imperfect specimens which we possess of it, to frame a contrasting diagnosis for it. Antidaphne we regard as truly Loranthaceous. We agree with Baillon in separating Myzodendron on account of its placentation. It is one of the connecting links which serve to bind the groups of Loranthaceae and Santalaceae very closely together.

8. Sur l'organogénie florale du Pleurandra.

9. Species Euphorbiacearum. A. Euphorbiacées Africaines. An enumeration chiefly of Cape species, including the genera Euphorbia, Cluytia, Acalypha, &c. Many of the distribution numbers of Drège, Zeyher, and others, are quoted.

10. Sur le Mercurialis alternifolia, et sur les limites du genre Mercuriale. M. Baillon proposes to modify the diagnosis of Mercurialis so as to include many species referred to Linozostis, Trismegista, Claoxylon, Erythrococca, Micrococca, Adenocline, Tragia, Acalypha, and Adelia, by authors.

11. Note sur les affinités du Rhodoleia. The ovary is said to be unilocular, with two parietal placentas, and so M. Baillon concludes it ought to be ranked amongst the Saxifrages.

12. Revue du Groupe des Vérbenacées (par M. Bocquillon). The classificatory portion of a memoir, part of which appeared in a previous volume.

13. Sur les ovules des Beaufortia. M. Baillon finds the cells of the ovary are at first 3-ovulate.

14. De l'influence de l'obscurité sur la Végétation. (Par M. Emery.) Referring to experiments, yet incomplete, instituted with a view to ascertain whether the formation of adventitious roots is attributable to deficient light, or to excess of moisture, &c. The development of the leaves and flowers of a hyacinth is described, the bulb of which was planted upside down, with its apex plunged in water.

15. Monographie du Groupe des Chloranthacées. (Par M. Cordemoy.) Based upon material much too imperfect. Sarcandra,

which depends upon characters, shown by Brown to be inconstant, viz. the suppression of the two lateral stamens, is maintained. Dr. Asa Gray's writings, in which reference is made to E. Asiatic Chloranthaceæ, are unknown to the author. His genus Saintlegeria is clearly the Tricercandra of Gray, described in the Botany of Perry's Japan Expedition' (ii. 318); S. gracilis being, no doubt, T. Fortunei of Gray.

At

Another Tricercandra is Chloranthus mandshuricus of Ruprecht. page 301 there must be some mistake or omission of a paragraph or page of MS., Ascarina polystachya, Forst., being quoted as a synonym of Sarcandra chloranthoides, and the generic character of Ascarina wholly omitted.

16. Description d'une Primulacée à fleurs monstrueuses. An interesting case of 'median prolification' in Lysimachia Ephemerum. No reference is made to its bearing upon the question of ovular homology, or to the views of Caspary as to the nature of the ovary of Primulaceæ.

17. Observations sur les affinités du genre Barbeuia. Another of Thouars' imperfectly known and ill described Madagascar plants. Baillon refers it to Phytolaccaces, instituting a new section Barbevieæ, characterised by a two-celled ovary, each cell being uniovulate : numerous stamens and hermaphrodite flowers.

18. Remarques sur l'organisation florale de quelques Bruniacées et sur les affinités du genre Grubbia.

19. Sur le Bosqueia, genre inédit de la Famille des Artocarpées. Yet another hitherto dubious plant of Thouars, from Madagascar. 20. Organogénie florale des Martynia. Studied with a view to settling the affinity of the genus.

21. Note sur l'albumen et l'arille des Hedychium. The albumen is said to be made up of elongated hair-like cells, converging upon the embryo, and containing starch granules. Baillon compares these cells with the pulp-cells of the orange. We apprehend, however, that his observations are not accurate, and that the constrictions to which he refers in the 'hair-like cells,' indicate points of junction of distinct cells. If we mistake not, such is the structure of the perisperm of various Scitamineae, and, less distinctly, of the Peppers, in both of which groups we have a double albumen, i. e., both endosperm and perisperm. M. Baillon would do well to examine seeds in various stages of development.

22. Sur les Fleurs monstrueuses de Sinapis arvensis. In one

flower, the placenta bore tiny flowers, instead of ovules. M. Baillon says: "Il est bien certain qu'on pourrait tirer, des faits * * * “énumerés, un bon nombre de conséquences contradictoires."

he adds, further, "Il y a peut-être, enfin, des botanistes qui ne vou"draient tirer de cette anomalie aucune conséquence. Ceux-là auraient "peut-être raison de dire:-Cette monstruosité prouve seulement "qu'il y a des monstruosités qui ne prouvent rien."

23. Première étude sur les Mappiées (Icacinées). M. Baillon agrees with Mr. Miers in approximating Icacineae and Ilicineae; going, indeed, further, and proposing to unite these groups in the same Family. We cannot at present discuss M. Baillon's opinions upon these plants at length.

LIII.-DE BARY ON VEGETABLE PARASITES.

RECHERCHES SUR LE DÉVELOPPEMENT DE QUELQUES CHAMPIGNONS PARASITES. MEMOIRE POUR SERVIR DE RÉPONSE A UNE QUESTION PROPOSÉ PAR L'ACADEMIE DES SCIENCES EN 1861, ET POUR SERVIR DE SUPPLEMENT AUX TRAVAUX SUR LA QUESTION

DES GENERATIONS DITES SPONTANÉES. Par M. A. De Bary, Professeur de Botanique, a l'Université de Fribourg en Brisgau. Annales des Sciences, IV. Serie, t. xx.

THE author states that his object in the present Memoir is to contribute to the solution of the following inquiries, viz., What is the origin of Parasitic Fungi? How do they reach the spots where their fruit is formed? What relation exists between them and the morbid condition of the organism they inhabit? or, as he puts it in other words, Can it be proved that in a diseased plant the mycelium of the endophyte grows from the diseased substance itself, and not from a germ of an endophyte of the same species; or do the observed facts explain the appearance of an endophyte by the development of its germs within the plant which bears it? The observations detailed in this paper are divided into three parts. First, those relating to Cystopus; secondly, those relating to Peronospora; and, thirdly, those on the family of the Uredineæ.

First, with regard to Cystopus:—the Cystopus is the occurrence of organs, discovered by Tulasne in Peronospora.*

principal fact observed in similar to those long since These oogonia, as they are

* Comptes rendus de l'Acad. des Sc. t. xxxviii. 26 June, 1854.

called, are concealed in the parenchyma of the nutrient plant, and are only visible externally by the brown colour of the surrounding tissue. Ramifications from the mycelium attach themselves to these oogonia, which ramifications are supposed to perform the functions of antheridia or male organs. Usually only one antheridium is attached to an oogonium. These two organs having completed their growth simultaneously, the granules contained in the oogonium accumulate in its centre, and form what Pringsheim has called the Befruchtungskugel. The antheridium then emits a tube, which is is supposed, by contact, to impregnate the contents of the oogonium. It is possible that this view may be correct, although we think that the process in other cryptogamic plants where impregnation has been observed, would certainly lead to the supposition that the presence of spermatozoa is essential for fecundation. If (as all observation tends to show) in ferns, mosses, and algæ, impregnation is effected only by spermatozoa, it seems hardly probable that fungi alone should exhibit the phænogamic feature of impregnation by contact of a tube. The author states that the oospores of Cystopus produce zoospores much in the same manner as the ordinary spores (or conidia as he calls them), and there seems to be no difference in the zoospores of the two organs. The latter are active for three or four hours; they then lose their cilia and power of motion, assume a cellulose covering, and germinate by the emission of a filament. These filaments are said to enter readily by the stomates of the leaves and stem of the nutrient plant, but in those situations they develop no further. It is only those germ-filaments which enter the stomates of cotyledons, which produce a mycelium. We think it may fairly be said that this latter observation requires confirmation. The excessive difficulty of tracing a mycelium to or from its origin in the germ-filament of a particular zoospore will justify some amount of scepticism upon the subject.

There is one other singular fact with regard to Cystopus which may be mentioned here. Dr. De Bary has noticed that of the plants upon which Cystopus occurs, some are more favourable to its development than others. For instance, Cystopus cubicus, although very common upon the Tragopogons, rarely if ever produces its sexual organs upon those plants, whilst in the cultivated Scorzonera these organs almost always accompany the conidia.

We now come to Peronospora. In this genus the development of the antheridia oogonia and oospores is said to be the same as in

Cystopus. If Dr. De Bary is right, Dr. Caspary has mistaken the young oogonia for a distinct kind of fruit; but it is hardly probable that so experienced an observer should not have been able to distinguish between the granular contents of the young oogonium and true sporidia.

In speaking of the germination of the conidia of Peronospora, the author notices that it varies in different species. In most conidia the summit is rounded, the thickness of the membrane is uniform, and they germinate (by the emission of a filament like other spores) at any point. But in P. ganglioniformis the conidia are papillate, and germination takes place only at the papilla. In P. densa and P. macrocarpa the protoplasm escapes and forms a cellulose membrane, and the globule thus formed emits a thick germ-filament from the point opposite to the opening of the conidium. Occasionally, but rarely, these two species germinate by terminal or lateral tubes, like the other species, with non-papillate conidia. In P. umbelliferarum and P. infestans the conidia, when placed in water, produce zoospores, which escape one by one through a cavity formed in the place of the papilla, which disappears. These zoospores soon become quiescent, assume a cellulose membrane, and emit a germfilament. Dr. De Bary remarks that in P. infestans the absence of light favours the production of zoospores, which, he says, are never produced when the conidia are exposed to full sunshine. This is a singular fact, as being directly contrary to what (in a former part of this Review) has been observed in the case of Reticularia,* and we believe that, as a general rule, full exposure to light is favourable, if not essential, to the production of zoospores.

A second mode of germination occurs in P. infestans, viz., the production by the conidia of secondary conidia, and by these secondary conidia of tertiary conidia, precisely resembling the original. These secondary and tertiary conidia, when plunged in water, produced zoospores.

A third mode of germination in the same species is by the production of a germ-filament, as in the species with round spores. This germination Dr. De Bary considers abnormal; but it is difficult to acquiesce in this view, grounded as it appears to be only on the fact of these germ-filaments having been produced in sowings in which the greater part of the conidia produced zoospores. It should be re

*Vol. ii. p. 406.