716 Commencing with British herbaria, the collection of the Royal Herbarium at Kew, generally acknowledged to be at onco the most extensive and the best preserved and most orderly in the world, comprises some 100,000 species, many of them represented by numerous specimens. It is arranged, for easy reference, in cases situate between the windows of the building containing it, the atmosphere of which is kept dry by means of hot-water pipes. In the intervening spaces are tables for the purpose of study, which is further facilitated by the presence in the same building of a large and valuable botanical library. Next in importance is the herbarium of the British Museum, which comprises assemblages of specimens gathered by numerous eminent botanists. The collection of Dillenius is deposited at Oxford, and that of the late Professor Harvey at Trinity College, Dublin. The original herbarium of Linnaeus is in the possession of the Linnean Society of London. With the more important British herbaria are to be ranked also those of Cambridge and Edinburgh. The collections of Jussieu and St Hilaire are included in the large herbarium of the Jardin cles Plantes at Paris, and in the same city is the extensive private collection of Dr Cosson. At Geneva are three large collections, De Canclolle s, containing the typical specimens of the Pro- dromus, Delessert s fine series, and Boissier s Mediterranean and Oriental plants. The university of Gottingen has had bequeathed to it the largest collection (exceeding 40,000 specimens) ever made by a single individual that of the late Professor Grisebach. At the herbarium in Brussels ara the specimens obtained by the traveller Martius, the majority of which formed the groundwork of his Flora Bmsiliensis. Other national herbaria sufficiently extensive to subserve the requirements of the systematic botanist exist at Berlin, St Petersburg, Vienna, Leyden, Stockholm, Upsala, Copenhagen, and Florence. Of those in the United States of America, the chief, formed by Asa Gray, is the property of Harvard university ; others are to be seen at Yale and Columbia colleges and at New York and Michigan universities. The herbarium at Melbourne, Australia, under Baron Miiller, has attained large propor tions ; and that of the Botanical Garden of Calcutta is noteworthy as the repository of numerous specimens de scribed by writers on Indian botany. Specimens of flowering plants and vascular cryptogams are, at Kew, generally mounted on sheets of stout smooth paper, of uniform quality, and in most cases 17 inches long by 11 inches broad; the palms and their allies, however, and some ferns, require a size of 22 by 14 inches. The tough but flexible coarse grey paper (German, Fliesspapier), upon which on the Continent specimens are commonly fixed by gummed strips of the same, is less hygroscopic than ordinary cartridge paper, but has the disadvantage of affording harbourage in the inequalities of its surface to a minute insect, Atropos puhatori<i, L., which commits great havoc in damp specimens, and which, even if noticed, cannot be dislodged without difficulty. The majority of plant specimens are most suitably fastened on paper by a mixture of equal parts of gum tragacanth and gum arable mada into a thick paste with water. Rigid leathery leaves are affixed by means of glue, or, if they present too smooth a surface, by stitching at their edges. Where, as in private herbaria, the specimens are not liable to be handled with great frequency, a stitch here and there round the stem, tied at the back of the sheet, or slips of paper passed over the stem through two slits in the sheet and attached with guni to its back, or simply strips of gummed paper lail across the stem, may be resorted to. A new adhesive substance, a kind of fish glue, has lately come into use for this purpose, and is highly spoken of. To preserve from insects, the plants, after mounting, are brushed over with a liquid formed by the solution of Ib each of corrosive sublimate and carbolic acid in 1 gallon of methylated spirits. They are then laid out to dry on shelves made of a network of stout galvanized iron wire. After this are written, usually in the right-hand corner of the sheet, or on a label there affixed, the designation of each species, the date and place of gathering, and the name of the collector. Information as to economical or medicinal properties may either be added thereto, or mentioned on the back of the sheet. It is especially important to attach to the name of the plant the initials or abbreviated name of the author by whom it was first described ; e.g , the words Ulva lactuca alone, might signify either of two distinct plants, the one described by Linnceus, the other by Agarclh. When the generic name has been altered, but the specific name has been retained, the name of the original describer of the plant is placed in brackets before the name of the later author. Thus the Enteromorpha Grevillei of Thuret, having been renamed, has become Ulva Grevillei (Thur.), Le Jolis. The value of specimens in private herbaria is greatly enhanced by briefly stating on the lower left-hand corner of the sheet the characters that distinguish it from the plants most nearly resembling it. Other particulars as to habit, local abundance, soil, and claim to be indigenous may be written on the back of the sheet, or on a slip of writing paper attached to its edge. It is convenient to place in a small envelope gummed to an upper corner of the sheet any flowers, seeds, or leaves needed for dissection or microscopical examination, especially where from the fixation of the specimen it is impossible to examine the leaves for oil- receptacles, and where seed is apt to escape from ripe capsules and be lost. The addition of a careful dissection of a flower greatly increases the value of the specimen. To ensure that all shall lie evenly in the herbarium the plants should be made to occupy as far as possible alter nately the right and left sides of their respective sheets. The species of each genus are then arranged either systematically or alphabetically in separate covers of stout, usually light brown paper, or, if the genus be large, in several covers with the name of the genus clearly indi cated in the lower left-hand corner of each, and opposite it the names or reference numbers of the species. Unde termined species are relegated to the end of the genus. Thus prepared, the specimens are placed on shelves or movable trays, at intervals of about 6 inches, in an air-tight cupboard, on the inner side of the door of which, as a special protection against insects, is suspended a muslin bag containing a piece of camphor. The systematic arrangement varies in different herbaria. The works usually followed are for dicotyledons, De Candolle s Prodromits, and Endlicher s or Bentham and Hooker s Genera Plantarum; for monocotyledons, Kunth s JEnumeratio ; for ferns, Hooker and Baker s Synopsis Fili cum; for mosses, Miiller s Synopsis Mitscorum Frondo- for algte, Kiitzing s Phycologia generalis ; for hepaticse, Gottschie, Lindcnberg, and Nees ab Esenbeck s Synopsis Ilepaticarum; and for other groups of cryptogamic plants the treatises of various authors scattered through numerous scientific publications. In certain herbaria, as in those of Boissier and Delessert at Geneva, the authority of the Prodromus is accepted only in the absence of any more recent treatise, or of a complete monograph on a family. For the members of large genera, e.y., Piper and Ficus, since the number of cosmopolitan or very widely distributed species is comparatively few, a geographical grouping is found specially convenient by those who are constantly receiving parcels of plants from known foreign source?. The ordinary systematic arrangement possesses
the great advantage, in the case of large genera, of readily