The good qualities of Erard's action, completed in 1821, the germ of which will be found in the later Cristofori, are not, however, due to repetition capability, but to other causes, chiefly, I will say, to counterpoise. The radical defect of repetition is that the repeated note can never have the tone-value of the first; it depends upon the mechanical contrivance, rather than the finder of the player, which is directly indispensable to the production of satisfactory tone. When the sensibility of the player's touch is lost in the mechanical action, the corresponding sensibility of the tone suffers; the resonance is not, somehow or other, sympathetically excited.

Erard rediscovered an upward bearing, which had been accomplished by Cristofori a hundred years before, in 1808. A down-bearing bridge to the wrest-plank, with hammers striking upward, are clearly not in relation; the tendency of the hammer must be, if there is much force used, to lift the string from its bearing, to the detriment of the tone. Erard reversed the direction of the bearing of the front bridge, substituting for a long, pinned, wooden bridge, as many little brass bridges as there were notes. The strings passing through holes bored through the little bridges, called agraffes, or studs, turned upward toward the wrest-pin. By this the string was forced against its rest instead of off it. It is obvious that the merit of this invention would in time make its use general. A variety of it was the long brass bridge, specially used in the treble on account of the pleasant musical-box like tone its vibration encouraged. Of late years another upward bearing has found favor in America and on the Continent, the Capo d'Astro bar of M. Bord, which exerts a pressure upon the strings at the bearing point.

About the year 1820, great changes and improvements were made in the grand pianoforte both externally and in the instrument. The harpsichord boxed up front gave way to the cylinder front, invented by Henry Pape, a clever German pianoforte-maker who bad settled in Paris. Who put the pedals upon the familiar lyre I have not been able to learn. It would be in the Empire time, when a classical taste was predominant. But the greatest change was from a wooden resisting structure to one in which iron should play an important part. The invention belongs to this country, and is due to a tuner named William Allen, a young Scotchman, who was in Stodart's employ. With the assistance of the foreman, Thom, the invention was completed, and a patent was taken out, dated the 15th of January, 1820, in which Thom was a partner. The patent was, however, at once secured by the Stodarts, their employers. The object of the patent was a combination of metal tubes with metal plates, the metallic tubes extending from the plates which were attached to the string-block to the wrest-plank. The metal plates now held the hitch-pins, to which the farther ends of the strings were fixed, and the force of the tension was, in a great measure, thrown upon the tubes.

The tubes were a mistake; they were of iron over the steel strings, and brass over the brass and spun strings, the idea being that of the compensation of tuning when affected by atmospheric change, also a mistake. However, the tubes were guaranteed by stout wooden bars crossing them at right angles. At once a great advance was made in the possibility of using heavier strings, and the great merit of the invention was everywhere recognized.

James Broadwood was one of the first to see the importance of the invention, if it were transformed into a stable principle. He had tried iron tension bars in past years, but without success. It was now due to his firm to introduce a fixed stringed plate, instead of plates intended to shift, and in a few years to combine this plate with four solid tension bars, for which combination he, in 1827, took out a patent, claiming as the motive for the patent the string-plate; the manner of fixing the hitch-pins upon it, the fourth tension bar, which crossed the instrument about the middle of the scale, and the fastening of that bar to the wooden brace below, now abutting against the belly-rail, the attachment being effected by a bolt passing through a hole cut in the sound-board.

This construction of grand pianoforte soon became generally adopted in England and France. Messrs. Erard, who appear to have had their own adaptation of tension bars, introduced the harmonic bar in 1838. This, a short bar of gun metal, was placed upon the wrest-plank immediately above the bearings of the treble, and consolidated the plank by screws tapped into it of alternate pressure and drawing power. In the original invention a third screw pressed upon the bridge. By this bar a very light, ringing treble tone was gained. This was followed by a long harmonic bar extending above the whole length of the wrest-plank, which it defends from any tendency to rise, by downward pressure obtained by screws. During 1840-50, as many as five and even six tension bars were used in grand pianofortes, to meet the ever increasing strain of thicker stringing. The bars were strutted against a metal edging to the wrest-plank, while the ends were prolonged forward until they abutted against its solid mass on the key-board side of the tuning-pins. The space required for fixing them cramped the scale, while the strings were divided into separate batches between them.

It was also difficult to so adjust each bar that it should bear its proportionate share of the tension; an obvious cause of inequality.

Toward the end of this period a new direction was taken by Mr. Henry Fowler Broadwood, by the introduction of an iron-framed pianoforte, in which the bars should be reduced in number, and with the bars the steel arches, as they were still called, although they were no longer arches but struts.

In a grand pianoforte, made in 1847, Mr. Broadwood succeeded in producing an instrument of the largest size, practically depending upon iron alone. Two tension bars sufficed, neither of them breaking into the scale: the first, nearly straight, being almost parallel with the lowest bass string; the second, presenting the new feature of a diagonal bar crossed from the bass corner to the string-plate, with its thrust at an angle to the strings.