and its comparative cheapness gives it an important place among the products of textile industry. It is used very largely in mixed fabrics, as well as for the cheaper ribbons, velvets, hat plush, and for many other silk woven fabrics, as also in the hosiery and glove trades and for sewing, knitting, and embroidering yarns. Silk Weighting.-Into the dyeing of silk it is not here necessary to enter, except in so far as concerns a nefarious practice, carried on in dye-houses, which has exercised a most detrimental influence on the silk trade. Silk, we have seen, loses about one-fourth of its weight in scouring. To obviate that loss it has long been the practice to dye some dark silks "in the gum," the dye combining in these cases with the gum or gelatinous coating, and such silks are known as "souples." Both in the gum and in the boiled-off state silk has the peculiar property of imbibing certain metallic salts largely and combining very firmly with them, the fibre remaining to external appearance undiminished in strength and lustre, but much added to in size and weight. Silk in the gum, it is found, absorbs these salts more freely than boiled-off; so to use it for weighting there are these great inducements--a saving of the costly and tedious boiling-off, a saving of the 25 per cent. weight which would have disappeared in boiling, and a surface on which much greater sophistication can be practised than on scoured silk. In dyeing a silk black a certain amount of weight must be added; and the common practice in former times was to make up on the silk what was lost in the scouring. Up to 1857 the utmost the dyer could add was "weight for weight," but an accidental discovery that year put dyers into the way of using tin salts in weighting with the result that they can now add 40 oz. per Ib to scoured silks, 120 oz. to souples, and as much as 150 oz. to spun silks, and yet call these compounds "silk." Not only so, but the use of tin salts, especially stannic chloride, SnCl, enables dyers to weight all colours the same as black. In his "Report on English Silk Industry" to the Royal Commission on Technical Instruction (1885) Mr Thomas Wardle of Leek says.— "Colours and white of all possible shades can very easily be imparted to this compound of silk and tin, and this method is becoming extensively used in Lyons. Thus weighting, which was until recently thought to apply only to black silks, and from which coloured silks were comparatively free, is now cheapening and deteriorating the latter in pretty much the same ratío as the former. Thus the proto- and per-salts of iron, as well as the proto- and per-salts of tin, including also a large variety of tannin, sumac, divi-divi, chestnut, valonia, the acacias (Areca Catechu and Acacia Catechu from India), from which are obtained cutch and gambier, &c., are no longer used solely as mordants or tinctorial matters, but mainly to serve the object of converting the silk into a greatly-expanded fibre, consisting of a conglomeration of more or less of these substances." Sugar also is employed to weight silk. On this adulterant Mr Wardle remarks: "With a solution of sugar, silk can have its weight augmented from 1 oz. to 3 oz. per Ib. I am not quite sure that this method of weighting was not first used by the throwsters, as sugar is known to have been used for adulterating and loading gum silk for a very long time, and then the idea was afterwards applied to silk after the dyeing operations. It is mach resorted to for weighting coloured has been found so cheap and easy of application. Bichloride of tin, hewing silks by dyers on the Continent, and, though a very clumsy method, no substitute chemical affinity for silk fibre, bids fair to extinguish the use of sugar, which, from its nygrometric qualities, has a tendency to ruin the silk to which it is applied, if great care be not taken to regulate the quantity. There is not the slightest use or excuse for the application of sugar, except to cheapen the silk by about 15 to 20 per cent." Wild Silk Dyeing.-Among the disadvantages under which the silks of the wild moths long laboured one of the most serious was the natural colour of the silks, and the extreme difficulty with which they took on dyes, specially the light and brilliant colours. For success in coping with this difficulty, as well as in dealing with the whole question of the cultivation and employment of wild silks, the unwearying patience and great skill of Mr Thomas Wardle of Leek deserve special mention here. The natural colour of tussur silk is a greyish fawn, and that shade it was found impossible to discharge by any of the ordinary bleaching agents, so as to obtain a basis for light and delicate dyes. Moreover, the chemical character of the tussur silk differs from that of the mulberry silk, and the fibre has much less affinity for tinctorial substances, which it takes up unevenly, requiring a large amount of dye-stuffs. After protracted experimenting Mr Wardle was able in 1873 to show a scries of tussurs well-dyed in all the darker shades of colour, but the lighter and bright blues, pinks, scarlets, &c., he could not produce. Subsequently the late M. Tessie du Motay found that the fawn colour of natural tussur could be discharged by solution of permanganate of potash, but the oxidizing action was so rapid and violent that it destroyed the fibre itself. Gentler means of oxidation have since been found for bleaching tussur to a fairly pale ground, but the dyeing of light colours cannot yet be said to be a commercial success. The silk of the eria or castor-oil worm (Attacus ricini) presents the same difficulties in dyeing as the common tussur. A portion of the eria cocoons are white, while the others are of a lively brown colour, and for the dyeing of light colours the latter require to undergo a bleaching process. The silk takes up colour with difficulty from a strong vat, and is consequently costly to dye. Moonga silk from Antheræa assama has generally a rather dark brown colour, but that appears to be much influenced by the leaves on which the worm feeds, the cocoons obtained on the champaca tree (Michelia champaca) giving a fine white fibre much valued in Assam. The dark colours are very difficult to bleach, but the silk itself takes dye-colours much more freely and evenly than either tussur or eria silk. Trade and Commerce. About the commencement of this century the chief silkproducing regions of the world were the Levant (including Broussa, Syria, and Persia), India, Italy, and France, the two first named sending the low-priced silk, the other two the fine qualities. Between 1840 and 1850, after the opening of trade with China, large quantities of silk were sent from the northern port of Shanghai, and afterwards also from the southern port of Canton. The export became important just at the time when disease in Europe had lessened the production on the Continent. This increased production of medium silk, and the growing demand for fine sorts, induced many of the cocoon-growers in the Levant to sell their cocoons to Europeans, who reeled them in Italian fashion under the name of "Patent Brutia," thus producing a very fine valuable silk. In 1857 commenced the importation of Japan silk, which became so fierce a competitor with Bengal silk as gradually to displace it in favour; and recently the native silk reeled in Bengal has almost ceased to be made, only the best European filatures, produced under the supervision of skilled Europeans, now coming forward. China and Japan, both of which contribute so largely to the supplies that appear in European and American statistics, only export their excess growth, silk weaving being carried on and native silk worn to an enormous extent in both countries. The other Asiatic exporting countries also maintain native silk manufactures which absorb no inconsiderable proportion of their raw material. The silk production of the world, including only the amount exported from these Oriental countries, amounts on an average to from 20,000,000 b to 25,000,000 ib yearly; but the crop is subject to great variations. The supply available for European consumption during recent years was thus stated, in bales of 100 b, by the Moniteur des Soies of Lyons, 25th July 1885: While these tables indicate remarkable fluctuation of | les Calais (tulles and passementerie), Paris, Nîmes, Tours, 'supply they show generally that Asiatic countries, besides Avignon, and Roubaix. Next to France in the extent supplying their own considerable demands, send to Europe and value of manufactures comes Germany, where the fully one half of the whole silk consumed in Western principal seat of the silk trade is at Crefeld, nearly one manufactures. China stands first as a silk-producing half of the whole production of the empire being manucountry, yielding about 35 per cent. of the entire supply; factured there. The looms of Crefeld and the district it the whole produce of Italy amounts to nearly the same controls numbered in 1881 about 33,000, and the trade proportion; the exports of Japan account for about 12 per was flourishing and expansive. The manufacture of union cent. of the annual supply; while in recent years France velvets is the special feature of the industry, about one and the Levant are credited with about equal proportions. half of the looms being devoted to that textile; but In the United Kingdom the trade in raw silks has been Crefeld controls also a large trade in union satins, and in a condition of decline for a considerable number of years, pure silk broad goods and ribbons of all kinds. The much of the Chinese and Eastern produce which formerly whole value of its trade amounted in 1881 to almost came to London now being unshipped at Marseilles, and £4,000,000, one-fourth of which found a market in sold in the Lyons market, which has become the leading England, and about a quarter of a million went to France. silk mart. But there is a very steady and continuous ex- The other principal centres of the silk trade, all in Rhenish pansion in the demand for waste silks and cocoons for the Prussia, are Viersen, Barmen, Elberfeld, and Mühlheim. spun silk trade. The following figures show the official Third on the list of Continental producers is Switzerland annual returns of silk imports since 1860, the date of the where Zurich takes the lead with broad goods (failles French commercial treaty, which exposed many branches armures, satins, serges, &c.), and Basel rivals St Étienne in of the trade to severe and fatal competition :the ribbon trade. The number of looms throughout the country is estimated at 40,000, of which 4000 are powerlooms. Italy-the early home of the silk trade, the land of the gorgeous velvets of Genoa and the damasks and brocades of medieval Sicily, Venice, and Florence—has fallen from its high estate, and now employs not more than 30,000 looms, the centre of greatest activity being at Como; but Genoa still makes velvets, and the brocades of Venice are not a thing of the past. In Austria the silk trade has found its principal development in Vienna and its immediate neighbourhood, the number of looms throughout the entire empire being estimated at from 15,000 to 20,000, of which 2000 are power-looms. In Russia there is, with a growing cultivation of raw silk, a considerable and increasing manufacture, the special feature of which is the weaving at Moscow of gold and silver tissues and brocades for sacerdotal use, and for traffic with Central Asia. : Silk 3 Years. Raw Silk. Knubs or Cwts. 1860-65 1865-70 1870-75 1875-80 1880 1881 2,904,580 1882 3,377,119 3,178,593 44,277 294,207 62,064 292,433 4,522,702 67,239 323,947 43,137,997 132,020 543,679 30,127,878 1883 1884 In the manufacture of silken fabrics France occupies the most important position among the nations. Not only is the whole of the raw silk produced in France worked up within the country, but a very considerable proportion of that imported from the Levant and from Asia passes into the hands of the French manufacturers. In all, between 8,000,000 and 9,000,000 lb of raw silk are on an average manufactured into various textures in France. Lyons is the headquarters of the trade, and, if the surrounding regions be included, employment is given to about 120,000 looms,-20,000 of which are driven by power, principally in the production of dress silks, plain and figured, and in other heavy silken fabrics, and at St Étienne and St Chamond in the ribbon trade. There are also important manufactures of silk at Calais, St Pierre The figures relating to Tsatlee comprise Re-reel, Hangchow, and Yuun-fa. Tho figures relating to Taysaam comprise Tussah. In the United Kingdom all the silk industries-these depending on spun silk alone excepted-have been in a depressed and declining condition ever since 1860. The principal silk manufacturing towns of England have been Coventry, Macclesfield, Congleton, Leck, Derby, London (Spitalfields), Manchester, Middleton, and Nottingham, and it is estimated that at the bost period not fewer than 150,000 looms found employment in the trade. In 1872 that number was reduced to 65,000, of which 12,500 were power-looms. Spitalfields in her best days (about 1825) kept 24,000 hand-looms occupied; now there are not more than 1200. Manchester once had about 20,000 looms weaving silk; now there are not 6000 so employed. When the French treaty of 1860 came into operation Coventry had about 9000 looms, principally employed in ribbon weaving; now not more than one-fourth of that number are in operation. The cause of several of these severe changes is to be found in the introduction of the factory system of working and the extension of power-loom weaving, which crushed out domestic weaving, the original form of the silk industry; but undoubtedly also the English manufacturers were beaten in the battle of free competition brought on by the French treaty. On the other hand, the remarkable development of the Silk and Silk Goods exported from the United Kingdom during the years 1860 to 1884 inclusive. comparatively new trade in spun silk goes far to compensate for the loss of the older trade, and has enabled the exports of silk manufactures from the country to be at least maintained and to show some signs of expansion. The spun-silk industry has chiefly developed in the Yorkshire and Lancashire textile centres, Bradford, Halifax, Rochdale, &c. But it is highly significant that, while the exports of British silk manufactures have not decreased, the imports in the meantime have shown a marked expansion; and unquestionably, although the use of silken goods has increased very greatly within twenty-five years, the expansion of native silk manufactures has not kept pace with that growth. Favoured by the operation of protective duties ranging from 50 to 60 per cent. ad valorem, the native manufacture of silk in the United States has been nursed into considerable activity and expansion, till now well-nigh one-half of the silken fabrics used in the country are of home manufacture. In 1860 the proportion of 12,000,000 12,000,000 5,000,000 tion was more especially due to the courses of public lectures which he delivered in the college and in various cities and towns of the United States. The happy combination of a graceful and interesting style with unwonted splendour of experimental illus tration gave these lectures an unprecedented popularity, and they exerted a powerful influence in awakening and developing a taste for scientific matters throughout the country. Besides the works already mentioned, Silliman published in 1808 an American 1866. SILLIMAN, BENJAMIN (1779–1864), American chemist and geologist, was born in 1779 at Trumbull (then called North Stratford), Connecticut. His father, Gold Selleck Silliman, was brigadier-general in the war of the revolution, and had also held important civil positions. The history of the family points to an Italian origin, but Daniel Silliman, the first to settle in the United States, came from edition of Henry's Chemistry, with notes, in 1827 an edition of Bakewell's Geology, Holland. Silliman received his early education at Fair- and in 1830 Elements of Chemistry, in two volumes. An account of his life, field, Connecticut, at that time the residence of his father's Prof. George P. Fisher, of Yale College, was published in two volumes in family, and in 1792 he entered Yale College, where he SILLIMAN, BENJAMIN (1816-1885), American chemist graduated in 1796. He then studied law, and was and physicist, son of the preceding, was born in 1816 at admitted to the bar in 1802, while a tutor in Yale College, New Haven, Connecticut, and educated at Yale College, to which position he had been appointed in 1799. In where he graduated in 1837. He then became assistant 1802 a professorship of chemistry and natural history was to his father in chemistry, mineralogy, and geology, workestablished in the college, and he was at once elected to filling in his laboratory at the college, and pursuing original it. He spent portions of. 1801 and 1802 in Philadelphia investigations. He began teaching in the laboratory soon in preparation for his work, and the year 1804 he spent in afterwards. The school thus informally established was Europe, chiefly in England and Scotland, where he attended shortly afterwards recognized by a formal act of the corthe lectures of Hope and Gregory, and also formed the poration of the college, and ultimately developed into the acquaintance of Davy, Wollaston, Brewster, Leslie, and Sheffield Scientific School of Yale College. In 1838 he other eminent men of science. As a result of this visit he became associate editor with his father of the American published A Journal of Travels in England, Holland, and Journal of Science and Arts, and he continued in the Scotland, and of Two Passages over the Atlantic in the years editorship of the journal until the close of his life, Prof. 1805 and 1806 (2 vols., 1810), which had a marked success. J. D. Dana (his brother in-law) having joined him in In 1813 he began service with the medical department of 1846. In the winter of 1845-46 he gave a course of Yale College as professor of chemistry and pharmacy, and lectures on agricultural chemistry in New Orleans, which continued to give instruction there for many years. In is believed to have been the first course of lectures upon 1818 he founded the American Journal of Science and that subject ever given in the United States. In 1849 he Arts, a periodical devoted to the physical sciences, which was appointed professor of medical chemistry and toxicohas been, and is, the most important American scientific logy in the medical department of Louisville university, serial. In 1851 he made a second journey to Europe, of Louisville, Kentucky, which position he held for five years. which he likewise published an account in two volumes, In 1854 he succeeded his father as professor of chemistry, edited by his son, who had accompanied him. In 1853 he and continued to give instruction in this science, first in became professor emeritus, but he continued to lecture for the academical and afterwards in the medical department a year or two longer. His closing years were quietly spent of Yale College, until his death in 1885. In 1853 he was in unabated mental activity at New Haven until his death connected with the exhibition at the Crystal Palace in New in 1864. Though devoted to scientific pursuits, he inter- York, having charge of the departments of chemistry, ested himself in the public movements of the time. geology, and mineralogy. As a result of this work he edited a large quarto volume, The World of Science, Art, and Industry (1853), followed in 1854 by The Progress of Science and Mechanism. He also published in 1846 First Principles of Chemistry, a text-book which had a wide sale and passed through three editions. In 1858 he published a manual of physics entitled First Principles of Physics or Natural Philosophy (2d ed. 1861). In 1864 and again in 1867 and 1872 Silliman visited California, being engaged in professional work connected with various mines and in mineralogical and geological explorations. Still later he made several visits to the mining regions of the western States and Territories, and the results of his observations formed the subjects of numerous scientific papers. In 1874, the centennial anniVersary of Priestley's discovery of oxygen, he delivered at Northumberland, Pa., where Priestley had resided during the later years of his life, an historical address on "Amer One of Silliman's earliest scientific publications was an account of the famous meteorite which fell in Weston, Conn., December 14, 1807. This account, which excited great public interest in the country, was reproduced abroad, and was read before the Royal Society of London, and also before the French Academy. Among his other scientific labours may be mentioned his experiments apon the fusibility of various substances in the flame of the com Dound blowpipe of Hare, then a novelty in science, and upon the vaporization and transference of the carbon in the voltaic arc Crom the positive to the negative pole, which he was the first to >bserve. He also repeated the experiment by which Gay-Lussac ad separated potassium from its hydrate, and obtained the element a its metallic form, doubtless for the first time in the United tes. Other professional labours were an exploration of the coal formations of Pennsylvania in 1830, and an examination of the gold mines of Virginia in 1836. In 1832 and 1833, by appointment of the United States Goverument, he made a scientific investigation of the culture aud manufacture of sugar, embodying his results in a voluminous report published by the Though Silliman published a large number of scientific papers upon chemical and geological subjects, his reputa Government. ican Contributions to Chemistry," which he afterwards expanded into a considerable volume. SILO. A modern silo is a pit or erection in which green crops are preserved in an undried condition for fodder. The term is derived from the Greek após (Lat. sirus), a pit for holding grain. It is only of recent years that ensilage, i.e., the preservation of green food for cattle by partial fermentation in silos, has become an important feature in agricultural economy. In various parts of Germany a method of preserving green fodder precisely similar to that used in the case of sauerkraut (see vol. iv. p. 618) has prevailed for upwards of a century. Special attention was first directed to the practice of ensilage by a French agriculturist, M. Auguste Goffart of Sologne near Orleans, who in 1877 published a work (Manuel de la Culture et de l'Ensilage des Mais et autres Fourrages Verts) detailing the experiences of many years in preserving green crops in silos. An English translation of M. Goffart's book by Mr J. B. Brown was published in New York in 1879, and, as various experiments had been previously made in the United States in the way of preserving green crops in pits, M. Goffart's experiof American dairy farming proved eminently suitable for the ensiling of green maize fodder; and the success of the method was soon indisputably demonstrated among the New England farmers. The favourable results obtained ence attracted considerable attention. The conditions in America led to much discussion and to the extensive introduction of the system in the United Kingdom, where, with different conditions, success has been more qualified, but still highly encouraging. It has been abundantly proved that ensilage forms a wholesome and nutritious food for cattle. It can be substituted for root crops with advantage, because it is succulent and digestible; milk resulting from it is good in quality and taste; it can be secured largely irrespective of weather; it carries over grass from the period of great abundance and waste to times when none would otherwise be available; and a larger number of cattle can be supported on a given area by the use of ensilage than is possible by the use of green crops. A silo should have a depth of at least 15 feet, and may either be a pit or a building above ground, provided it is water-tight and, as far as possible, air-tight. The crops suitable for ensilage are the ordinary grasses, clovers, lucerne, vetches, oats, rye, and maize; but various weeds may also be stored in silos with good results, notably spurrey, (Spergula arvensis), a most troublesome plant in poor light soils. As a rule the crop should be mown when in full flower, and deposited in the silo on the day of its cutting. Fair dry weather is not essential; but it is found that when moisture, natural and extraneous, exceeds 75 per cent. of the whole, good results are not obtained. The material is spread in uniform layers over the floor of the silo, and closely packed and trodden down. If possible, not more than a foot or thereby should be added daily, so as to allow the mass to settle down closely, and to heat uniformly throughout. When the silo is quite filled a layer of straw or some other dry porous substance is spread over the surface, then it is covered with boards, and a pressure of not less than 100 fb per square foot is applied by weighting or other mechanical means. A silo thus contains, to begin with, a mass of living vegetable cells surrounded with a minimum of oxygen. The activity of the cells continuing, oxygen is absorbed and carbonic acid evolved, and part of the starch of the plants is converted into sugar. In the atmosphere of carbonic acid thus created the acid ferments manifest their vitality, and acetic, lactic, and butyric acids are developed at the expense of the starch and sugar. These chemical changes are accompanied with an evolution of heat, and the temperature of the mass rises, till, when it attains 122° Fahr., the action of the ferments is arrested. Should the heat rise to 150° the vitality of the vegetable cells themselves is destroyed; and also when the available oxygen is exhausted chemical change ceases and sweet silage is produced. When from excess of moisture or other cause the temperature of the silo does not reach 122° Fahr., the acid ferments are not killed, and they go on evolving chiefly acetic and lactic acids, the results being sour silage. These ferments. requiring nitrogen for their existence, act on the nitrogenous constituents of the plants, rendering the albuminoids partly soluble, evolving peptones, and by further splitting up producing amides, urea, and ammonia. The formation and loss than is incident to sweet silage; and in extreme production of sour silage is accompanied by much greater transaction the material acquires a most disagreeable odour. There is, however, no sharp line of distinction between the two, and both varieties are eaten freely by stock. Frequently a considerable loss occurs around the edges, and at other points where air gets access to the mass, by mildewing. See Report of Select Committee. flourishing in the present epoch, and represented by a SILURIDÆ, a large family of freshwater Fishes, great variety of forms in all the tropical and temperate great variety of forms in all the tropical and temperate regions, many of them reaching back into the Tertiary age. The principal characters of this family (termed a "suborder" by some), its position in the system, its geographical distribution, and some of the most remarkhave been already sufficiently noticed under ICHTHYOLOGY, able points in the structure and life-history of its members but we have here to notice more fully the sections into which it has been divided, and certain remarkable forms which were referred to nominally only in that article. specialization of certain portions at the expense of others, The modifications of the vertical fins, or rather the and the greater or less extent of the branchial aperture form excellent characters for subdividing the Siluroids. I. In the Silurida Homaloptera the vertical fins are exceedingly long, occupying nearly the whole extent of the embryonal fin, and in one genus (Heterobranchus) a great part of the dorsal portion retains its embryonic character, being a rayless adipose fin. All the Siluroids of this section belong to the fauna of the Old World and Australia. The rivers and lakes of tropical Africa harbour many species of the genera Clarias and Heterobranchus,-those of the Nile being known under the name of "Carmoot." One of the Nilotic species, Clarias macracanthus, occurs abundantly in the Lake of Galilee, and, being a long, scaleless, eel-like broad mouth, was certainly included among those which fish of black colour, with eight long barbels round its the Jews were forbidden to eat by the Mosaic law. These fish grow to a length of from 4 to 6 feet, and are eaten by the natives of tropical. Africa. II. In the Silurida Heteroptera the dorsal fin has almost or entirely disappeared; only its foremost portion and a small adipose remnant may be preserved; on the other hand the anal portion is retained in its whole extent. The gillmembranes remain separate and overlap the isthmus. This section likewise belongs to the fauna of the Old World, and includes, among many others, the species which has given the name to the whole family Silurus glanis, the "Wels". FIG. 1.-The "Wels" (Silurus gianis). of the Germans. It is the only representative of the family in Europe, and with the exception of the sturgeon, is the largest freshwater isn of the Continent. It was known to Aristotle, who described it under the name of Glanis. It inhabits more the central and eastern portions of Europe than the western, being absent in Italy, Greece, southern Switzerland, France, and those parts of Germany which are drained by the Rhine and its affluents. In general appearance it somewhat resembles the burbot. Its head is large and broad, its mouth wide, furnished with six barbeis, of which those of the upper jaw are very long. Both jaws and the palate are armed with broad bands of small closelyset teeth, which give the bones a rasp-like appearance. The eyes are exceedingly small The short body terminates in a long, compressed, muscular tail, and the whole fish is covered with a smooth, scaleless, slippery skin. Specimens of 4 and 5 feet in length, and of 50 to 80 b in weight, are of common occurrence. Its food consists chiefly of other bottom-feeding fishes, and in inland countries it is considered one of the better class of food fishes. Stories about children having been found in the stomach of very large individuals are probably inventions. III. The Silurida Anomalopteræ are a small section from tropical America, in which the dorsal and adipose fins are very short and belong to the caudal vertebral column, while the anal is very long, and the gill-membranes are entirely separate, overlapping the isthmus. Some IV. The Silurida Proteroptera are a section extremely numerous in species, and represented throughout the tropics. The dorsal fin consists of a short-rayed and an adipose portion, the former belonging to the abdominal vertebral column; the anal is always much shorter than the tail. The gill-membranes are not confluent with the skin of the isthmus; they have a free posterior margin. When a nasal barbel is present, it belongs to the posterior nostril. This section includes among many others the genus Bagrus, of which the "Bayad" (B. bayad) and "Docmac" (B. docmac) frequently come under the notice of travellers on the Nile; they grow to a length of 5 feet, and are eaten. Of the "Cat-Fishes" of North America (Amiurus), locally called "bull-heads" or "horned-pouts," with eight barbels, some twenty species are known. of them are valued as food, especially one which is abundant in the ponds of New England, and capable of easy introduction into other localities (4. nebulosus). Others which inhabit the great lakes (A. nigricans) and the Mississippi (A. ponderosus) often exceed the weight of 100 b. Platystoma and Pimelodus people the rivers and lakes of tropical America, and many of them are conspicuous in this fauna by the ornamentation of their body, by long spatulate snouts, and by their great size. The genus Arius is composed of the greatest number of species (about seventy), and has the widest distribution of all Siluroids, being represented in almost all tropical countries which are drained by large rivers. Some of the species enter salt water. They possess six barbels, and their head is extensively osseous on its upper surface; their dorsal and pectoral spines are generally developed into powerful weapons. Bagarius, one of the largest Siluroids of the rivers of India and Java, exceeding a length of 6 feet, differs from Arius in having eight barbels, and the head covered with skin. V. In the Silurida Stenobranchiæ the dorsal fin consists of an adipose portion and a short-rayed fin which belongs to the abdominal vertebral column, and, like the adipose fin, may be sometimes absent. The gill-membranes are confluent with the skin of the isthmus. The Siluroids belonging to this section are either South-American or African. Among the former we notice specially the genus Doras, which is distinguished by having a series of bony scutes along the middle of the side. The narrowness of their gill-openings appears to have developed in them a habit which has excited the attention of all naturalists who have visited the countries bordering upon the Atlantic rivers of tropical America, viz., the habit of travelling during seasons of drought from a piece of water about to dry up to ponds of greater capacity. These journeys are occasionally of such a length that the fish have to travel all night; they are so numerous that the Indians fill many baskets of them. Hancock supposes that the fish carry a small supply of water with them in their gill-cavity, which they can easily retain by closing their branchial apertures. The same naturalist adds that they make regular nests, in which they cover up their eggs with care and defend them,-male and female uniting in this parental duty until the eggs are hatched. Synodontis is FIG. 2.-Synodontis xiphías. an African genus and common in the Nile, where the various species are known by the name of "Shal." They frequently occur among the representations of animals left by the ancient Egyptians. The upper part of their head is protected by strong osseous scutes, and both the dorsal and pectoral fins are armed with powerful spines. Their mouth is small, surrounded by six barbels, which are more or less fringed with a membrane or with branched tentacles. Finally, the Electric. Cat- or Sheath-Fishes (Malapterurus) also belong to this section. Externally FIG. 3.-Malapterurus electricus. they are at once recognized by the absence of a rayed dorsal fin, of which only a rudiment remains as a small interneural spine concealed below the skin. The entire fish is covered with soft skin, an osseous defensive armour having become unnecessary in consequence of the development of a powerful electric apparatus, the strength of which, however, is exceeded by that of the electric eel and the large species of Torpedo. It has been noticed in vol. xii. p. 650. Three species have been described from rivers of tropical Africa, of which one (M. electricus) occurs in the Nile; it rarely reaches a length of 4 feet. VI. The section of Silurida Proteropodes contains small forms, some of which are of interest by the degree of specialization to which they have attained in one or the other direction. Many of them are completely mailed; but all have in common a short-rayed dorsal fin, with the ventrals below or rarely in front of it. Their gill-openings are reduced to a short slit; their pectorals and ventrals have assumed a horizontal position; and their vent is before, or not much behind, the middle of the length of the body. The first group of this section comprises alpine forms of the Andes, without any armature, and with a very broad and pendent lower lip. They have been referred to several genera (Stygogenes, Arges, Brontes, Astroplebus), but are collectively called "preñadillas" by the natives, who state that they live in subterranean craters within the bowels of the volcanoes of the Andes, and are ejected with streams of mud and water during eruptions. These fishes may, however, be found in surface waters at all times, and their appearance in great quantities in the low country during volcanic eruptions can be accounted for by numbers being killed by the sulphuretted gases which escape during an cruption and |