homogeneous wrought-iron plate, a successful
competitor was long in coming, though both
compound and steel plates early appeared in the
competition, as has already been stated. In
1S59 the Mersey Works in England submitted
2.5-inch plates made up of three layers, the outer ones of wrought iron welded to a middle one of
steel ; the welding was imperfect and the brittleness of the steel affected the outer layers ; moreover, they were constructed on a wrong principle,
as will be seen presently in connection with
capped projectiles. In 1801 compound and steel
plates wei'e tested in England ; while a long
series of tests with steel was carried on in
France from 1857 to 1801. The theoretical ad-
vantages of steel and compoiuid armor were
thoroughly understood at this time, but brittle-
ness defeated one and imperfect welding com-
bined with brittleness precluded the use of the
other. The preeminence of wrougiit-iron plate
continued practically undisputed until 1876. In
1875, the Italian Government called for test
plates of 22-inch solid armor from all the prin-
cipal manufacturers in the world. Four re-
sponded — Brown & Co. submitted two solid iron
plates; Cammell & Co. one solid plate and
one sanilwicli target; Marrell et Cie. pi'esented
the same; and Schneider et Cie. presented
for test two solid steel plates. The trials
took place at Spezia in 1870, and the re-
sult was overwhelmingly in favor of the steel
plates ; though these cracked more than the
others, the dilTerence in resistance to penetra-
tion was great. As a result of the Spezia test,
wrought-iron plates were fully discredited. In
England, where the prejudice against steel was
fostered to some extent by national jealousy of
a French manufacturer's success, compound ar-
mor was rapidly developed. This consisted of a
hard steel face welded to a wrought-iron back.
In the Wilson process (Cammell & Co.), molten steel is poured on a white-hot iron plate, tlius adding one-third to the thickness of the plate, whicii is afterwards rolled. In the Ellis process (lirown & Co.) a wrought-iron' and a rather thin, hard steel plate are separately made. The lower plate is placed in the furnace; iron bars are placed around the edge forming a berm, and the steel plate laid on the bars. After they are raised to a welding heat, molten steel is (loured into the space between the plates, welding all together. There are other sj'stems of cora- ])Oiuul plate manufacture, but these have at- tained the greatest eminence, if extent of use is a criterion, and they exemplify the type. The first really successful compound plates appeared in 1877, and from that date they competed with steel plates on apparenth- equal terms until 1880, when Scluieider brought out his first nickel-steel plate. The trial of this |ilate took place in July, but it was not until the next year that attention was fully drawn to the importance of the new metal. The United States naval authorities, having purchased in Europe a compound plate from Canunell & Co., a steel jdate and also one of nickel steel from Schneider et Cie., held a (Mimpetitive test at the naval proving-ground, Annapolis, in September, 1890. The compound plate was ignominiously defeated, both by the steel plate and by that of nickel steel. Indeed, the defeat was sn complete and convincing that it stopped the manufacture of compouml jilates at once. The Navy Department had already com- mitted itself to the manufacture of steel armor, and large contracts for it had been given out some years before. The wisdom of the decision in favor of steel had been questioned in Con- gress and in the public press, and the trial was instituted to convince the doubters ; the occa- sion also served to etl'ect a trial of a nickel-steel plate which had been acquired on account of the promising results obtained by Messrs. Schnei- der et Cie the year before. The important re- sults of this trial ha<i hardly appeared when a second trial of like revolutionary character took place. A 3-ear or two before, Mr. Harve}', a manufacturer of fine-grade tool steel, was at the Washington gun factory. In a conversation with him. Captain Folger, the superintendent of the naval giui factory, suggested the attempt to adapt the Harvey system of hardening tool steel to armor plate. After considering the matter, Mr. Harvey decided to make some experi- ments looking to surface hardening armor ; and the result was the Harvey process of surface carburization- and hardening. In this pi'ocess the plate is placed on the floor of a suitable furnace in a bed of refractory mate- rial, leaving the surface to be hardened upper- most. This is covered with carbonaceous mate- rial, which is rammed down upon it; over the carbon is put a layer of sand covered in with fire-brick. The temperature of the furnace is then raised to about the temperature of melting cast iron and kept so for several days, until the required additional carburization — usually about 1 per cent. — is effected. The plate is then removed, and, when cooled to a dull cherry red, is hardened by a water jet or immersion in run- ning water. The trial of the first plate made by the Harvey process took place February 14, 1891, and was very successful. The popularity of the new armor was immediate, and the cause is not far to seek, for in its construction it embodied the combined views of armor theorists.
Supporters of compound armor accepted the defeat of the year before with reservations. They still adhered to the idea of a hard face and soft back; and they were right in so doing. The ad- vocates of a single, practically homogeneous plate were also satisfied: they held, very properly, that the strength of iron armor plate must not lie solely in defeating the projectile by breaking it up. but in kee]iing it out by the toughness of the plate. A third class of critics — more or less in sympathy with the compound armor people, but not comidetely so, and declaring that no armor is fit to put on a ship which breaks up under attack and leaves the side bare after- wards, even if it stops that particular projectile, — were also satisfied. In England, the Harvey system was adopted almost at once, but notions of misplaced economy and the difticulty of work- ing the resulting metal caused the rejection of nickel in steel. But this view was held only a few years. The next improvement, in 1895, was effected by the Carnegie Company, which found that a reforging of plates after carburizing considerably improved the quality. Soon after the appearance of Harveyized armor, Krupp, Schneider, the Terni ^'orks, and several other European makers began experiments along the same lines. The ones mentioned alone produced any noteworthy results, and it is to be remarked that all three adopted the method of carburization by means of gas rich in carbon. Of these three
