to emphasize, in conclusion, the important position the Pteridophyta hold with regard to the existing flora.
Cultivation.—Numerous species of ferns, both temperate and tropical, are cultivated as valued ornamental plants. Species of the other groups are occasionally grown for scientific purposes in the larger botanic gardens, but their cultivation, which often presents special difficulties, need not be referred to here. While a number of ferns can be multiplied vegetatively, by buds formed on the leaves and in other ways, the regular mode of propagation is by sowing the spores shed from the ripe sporangia. The spores should be thinly sprinkled on the surface of the soil in well-drained pots, which should stand in saucers filled with water and be covered with glass plates. After the prothalli have attained some size and bear sexual organs the pots should be occasionally sunk in water so as to flood the prothalli for a few minutes and facilitate fertilization. The young plants developed on the prothalli should be carefully pricked out into other pans and later transferred to 3-in. pots. When the pots are fairly filled with roots the plants may be shifted into larger ones.
The best time for a general repotting of ferns is in spring, just before growth commences. Those with creeping rhizomes can be propagated by dividing these into well-rooted portions, and, if a number of crowns is formed, they can be divided at that season. In most cases this can be performed with little risk, but the Gleichenias, for example, must only be cut into large portions, as small divisions of the rhizomes are almost certain to die; in such cases, however, the points of the rhizomes can be led over and layered into small pots, several in succession, and allowed to remain unsevered from the parent plant until they become well rooted. In potting the well-established plants, and all those of considerable size, the soil should be used in a rough turfy state, not sifted but broken, and one-sixth of broken crooks or charcoal and as much sand as will insure free percolation should be mixed with it.
The stove ferns require a day temperature of 65° to 75°, but do not thrive in an excessively high or close dry atmosphere. They require only such shade as will shut out the direct rays of the sun, and, though abundant moisture must be supplied, the atmosphere should not be loaded with it. The water used should always be at or near the temperature of the house in which the plants are growing. Some ferns, as the different kinds of Gymnogrammae and Cheilanthes, prefer a drier atmosphere than others, and the former do not well bear a lower winter temperature than about 60° by night. Most other stove ferns, if dormant, will bear a temperature as low as 55° by night and 60° by day from November to February. About the end of the latter month the whole collection should be turned out of the pots and redrained or repotted into larger pots as required. This should take place before growth has commenced. Towards the end of March the night temperature may be raised to 60°, and the day temperature to 70° or 75°, the plants being shaded in bright weather. Such ferns as Gymnogrammas, which have their surface covered with golden or silver powder, and certain species of scaly surfaced Cheilanthes and Nothochlaena, as they cannot bear to have their fronds wetted, should never be syringed; but most other ferns may have a moderate sprinkling occasionally (not necessarily daily) and as the season advances sufficient air and light must be admitted.
Authorities.—Scott, Structural Botany: Flowerless Plants (London, 1896), Studies in Fossil Botany (Edinburgh, 1900);* Campbell, Mosses and Ferns (London, 1895);* Engler and Prantl, Die naturlichen Pflanzenfamilien (Theil i. Abth. 4.; Leipzig, 1898-1902); Bower, The Origin of a Land Flora (London, 1908); Goebel, Organography of Plants (Oxford, 1905); Hooker and Baker, Synopsis Filicum (London, 1874); Baker, Fern Allies (London, 1887); Christ, Die Fankräuter der Erde (Jena, 1897); Seward, Fossil Botany, vol. i. (Cambridge, 1898). In those works marked with an asterisk copious references to the recent literature of the subject will be found.
PTEROBRANCHIA, a zoological group established by Ray Lankester in 1877. It contained at that time the single genus Rhabdopleura, a minute animal dredged by Sars off the Lofoten Islands, and by Norman off the Shetlands. Rhabdopleura was at first regarded as an aberrant Polyzoon, but with the publication of the Challenger Report (Cephalodiscus) in 1887, it became clear that Cephalodiscus, the second genus now included in the order, had affinities in the direction of the Enteropneusta. The connexion of the Pterobranchia with the Polyzoa is in the highest degree questionable.
Rhabdopleura is no doubt of world-wide distribution, since it has been recorded in various localities from Greenland to South Australia, usually in water of not less than forty fathoms. Cephalodiscus, which for many years was known solely as the result of a single dredging by the “Challenger” from 245 fathoms in the Straits of Magellan, has recently been found in entirely different parts of the world, as for instance between Japan and Korea at 100 fathoms, at about half that depth off the south-east coast of Celebes, and between tide-marks on the coast of Borneo. It appears to be common in the neighbourhood of Cape Town, while the recent Antarctic expeditions have shown that it occurs in various localities from the Falkland Islands to the Antarctic circle. No less than twelve species, referred to three sub-genera (Demiothecia, Idiothecia, Orthoecus), have now been described; but it is at present uncertain whether more than a single species of Rhabdopleura is valid, although several specific names have been suggested for specimens from different localities.
Both genera are characterized by their habit of secreting a tubular gelatinoid investment, the “coenoecium,” composed of a number of superposed lamellae, doubtless the result of its intermittent secretion, mainly though perhaps not exclusively, by the proboscises of the zooids. In Rhabdopleura each zooid forms its own delicate tube composed of a characteristic series of distinct rings. In Cephalodiscus the coenoecium is more massive, and may contain a continuous irregular cavity in which the zooids live (Demiothecia), or may be secreted in such a way that each zooid has its own independent tube (Idiothecia, Orthoecus).
The zooids are a modification of the type of structure known in Balanoglossus, from which they differ principally in the following respects: (i.) The alimentary canal, instead of being straight, has a U-shaped flexure, the dorsal line between the mouth and the anus being short. (ii.) The proboscis (fig. 1, b), known as the “buccal shield,” is a large organ, strongly flattened in an antero-posterior direction, its ventral lobe usually concealing the mouth. (iii.) The collar is produced dorsally into arms (one pair in Rhabdopleura, four to eight pairs in Cephalodiscus), each of which bears numerous ciliated tentacles, the organs by which the microscopic food-particles are conveyed to the mouth. (iv.) The third division of the body, the metasome, is prolonged ventrally into a relatively enormous outgrowth containing the loop of the alimentary canal, beyond which projects a stalk (fig. 1, c), of a length varying with the state of contraction and perhaps with the species. (v.) The stalk gives rise to buds, by which the colonial habit is acquired. While in Rhabdopleura the buds remain in organic continuity with the parent, in Cephalodiscus they become free at an early stage, and the coenoecium accordingly contains a number of separate individuals. In the living Cephalodiscus a zooid can crawl by means of its proboscis over the gelatinous processes of the outer side of the coenoecium, a position which it can assume owing to the very great extensibility of the stalk, the proximal suctorial end of which remains attached to the inner surface of some part of the coenoecium (Andersson, 1907).
 | |
| (From a drawing by Professor McIntosh.) | |
|
Fig. 1.—Zooid of Cephalodiscus dodecalophus. | |
| a, a, | Buds. |
| b, | Proboscis. |
| c, | Stalk. |
| d, | Arms and tentacles. |
| e, | Ventral edge of proboscis. |
| f, | Its dorsal edge. |
In correspondence with the fundamental constitution of the zooid, each of the three segments has its own body-cavity separated from the others. The main proboscis-cavity (fig. 2, b.c.1) is unpaired, and opens to the exterior by the two proboscis pores (p.p.). It contains a closed vesicle regarded by Schepotieff as a right proboscis-cavity and in any case representing the pericardium of Balanoglossus, the glomerulus of which is also probably represented. The collar-cavity (b.c.2) is paired, although its ventral mesentery is not complete. It extends into the arms, which originate in the bud (fig. 1) as dorsal outgrowths of the collar. The ventral and lateral parts of the anterior
