"Extending around the inner margin of each sash is a groove 1, made of suitable width to receive the edges of the panes of glass & and of a depth desirably about equal to its width, except along the upper edge of the sash, where the groove takes the form of an open slot 2, into which the glass may be thrust considerably farther than the distance which it normally occupies when in place, Figs. 2 and 5.

"The glass G for each sash is provided in a plurality of panes in this, instance two-the aggregate width of which is made somewhat less than the extreme inner width of the sash from the bottom of the groove 1 on one

side to the bottom of the same groove on the other. This permits the glass to be inserted by proceeding in the manner indicated in Figs. 5 to 9, inclusive. The upper edge of the first pane is slipped angularly up through the slot 2, as shown in Fig. 5, and the whole pane is then swung into a vertical position and slipped down until its lower edge enters the groove 1 at the bottom of the sash. The pane then occupies a position such as is indicated in Fig. 6 and

can be slipped sidewise until one of its lateral edges enters the groove 1 at the side of the sash, as shown in Fig. 7. This leaves a space between its other lateral edge and the other side of the sash wide enough to receive the second pane, which is then inserted in the same manner as the first pane until it occupies the position shown in Fig. 8, whereupon it is moved sidewise away from the first pane until its lateral edge enters the groove 1 on the other side of the sash, as shown in Fig. 9. The glass is then secured in this position, as follows: Midway between its sides the sash is provided with a vertical bar 3, located in front of the groove 1, with its rear face in the same plane as the front edge of said groove or in line with the front surface of the inserted glass. This bar covers the adjacent edges of the two panes of glass on their front sides and affords a ledge against which these inner edges of the glass rest. Operating in conjunction with this middle sash bar 3 is a clamping bar 4, having forwardly projecting flanges 5, which enter between the panes, so as to prevent their lateral movement, and lateral flanges 6, which project back to the glass and hold it firmly against the front bar 3. This clamping bar is herein shown as secured in place by screws 7, which extend through the bar into screw plates 8 in the front bar 8 and by turning up which the glass may be held between the clamping bar and front plate as tightly as desired. This clamping bar is put in place after the panes have been inserted and moved laterally into their final positions, as shown in Fig. 9, and the tightening of its screws 7 completes the glazing operation.

"One great advantage of the construction described is that it enables the window to be glazed from the inside of the building and does not require any attention or manipulation from without, as is commonly necessary in fireproof windows of this character. Windows so constructed may, furthermore, be made of any desired width and arranged to contain as many panes of glass as may be necessary to glaze the desired width of the sash, without regard to the width of the panes, since it is only necessary to provide as many additional bars 3 and 4 as are needed in clamping the adjacent edges of each pair of panes. Obviously many changes may be made in the details of the construction shown without departure from the broad spirit of the invention claimed." There has been much discussion of the question whether Watson invented a "fireproof" window-complainant asserting that he did; defendant that he did not. Of course the problem of defining a fireproof window, the inclusion of features necessary, and the exclusion of those unnecessary, to constitute such a structure, is not before the court for solution. Common observation teaches us that, after all, structures are "fireproof" as a matter of degree only. The term expresses the quality of one structure to withstand the stress of fire relatively to, or by comparison with, those which are deemed flammable or not so capable to withstand it. Therefore, in the present case, it may be conceded that a metal sash window with wire glass is, acceptably, a "fireproof window," simply because the materials are noninflammable. But upon the issues here the question is of importance, because upon its answer may depend the scope of the art to which reference may be had in determining the quality of Watson's endeavors-whether he invented anything, or merely selected and adapted to his structure well-known features. Upon the specifications and drawings Watson's invention is thus embodied in the first of the four claims of the patent:

"In a window, the combination of a plurality of panes of glass, of a sash grooved to inclose said panes of glass, a bar fixed to and extending across the sash in front of the plane of the groove and past which the panes may be slid lateral portions of the groove, a clamping bar interposed between the panes when slid into the side grooves and bearing against their adjacent edges opposite to the fixed bar, and means for detachably securing the clamping bar in place, substantially as described."

The second claim differs only in narrowing the "means for detachably securing the clamping bar" to "clamping screws," while the third and fourth claims differ substantially from the first and second, in respectively incorporating a "metal" sash. The five elements of the combination are therefore: (1) The plurality of the panes of glass; (2) the grooved sash for their inclosure; (3) the bar fixed to and extending across the sash in front of the plane of groove,

past which the panes may be slid laterally to introduce or withdraw them; (4) the clamping bar interposed between the panes; (5) the means for securing the clamping bar.

In determining the scope of the art to which Watson must be held, there are several considerations which are, I believe, controlling, in support of the conclusion that the invention, as claimed, is not a "fireproof window": First, his statement that his invention relates "to improvements in windows," and "more particularly to fireproof windows of that class which have their frames made of sheet metal," with the object of providing "improved construction for securing the glass in place in windows of this character"; secondly, the fact that a "metal" sash appears as an element in claims 3 and 4 and not in claims 1 and 2, taken in connection with the substantial identity of the claims as to other elements, as already indicated, shows, no matter what may be said of their validity, that the claimed improvement relates to an extended art; third, the rejection by the Patent Office of a claim for a fireproof window, for which no substitute claim now appears in the patent.

It may be, as contended by plaintiff, that Watson had in view the adaptability of his structure to the use of wire glass through which the fireproof quality would be greatly strengthened and more nearly approach perfection; that any expectation of its utility in such combined use has been realized through the advent and extensive use of that kind of glass. But his real invention is neither more nor less than he has declared-an improvement in windows-and there is open for consideration the entire art relating to sash for windows or window structure. Therefore the references cited by defendant, hereafter referred to, are deemed relevant.

Now in taking this view, defendant refers us to numerous structures of the prior art, ranging from simple forms of wooden sash structures of doors and windows to rather complex forms of fireproof windows and of sash in use in sky windows. No reason is perceived why these references are not pertinent, or why the more fundamental art of glazing joinery should not be consulted to determine the fund of information presumably possessed by mechanics in that art, in order to answer the question respecting Watson's alleged invention. Therefore the Sheldon patent, though for a screen door disclosing the form and arrangement of grooves for panels, and Short's disclosure, in a wooden window sash, of a removable muntin, are not to be discarded, in considering Watson's structure as one not necessarily fireproof. But an examination of the Hayes, the Bickelhoupt, the Rendle, and the Coulson patents shows, in my judgment, that Watson really does no more than to select from them, respectively, elements now found in his claims. Certainly the muntin of Watson differs from Short only in the fact that in the one the rib is upon the fixed, in the other upon the removable, part. The grooves are alike, both in arrangement and shape, in Watson, Sheldon, and Hayes. Watson's feature of deepening one groove to receive the pane and to enable its being lifted, then lowered, into the opposite groove, is disclosed and expressly claimed by Hayes. With respect to the clamping bar and the means for securing it, and which is claimed "substantially as described," I am unable to find any difference in the disclosure shown by Watson in Figure 3, supra, and Figures 17 and 18, viz.:

and those made by Coulson, patent 429,375, and Rendle, No. 1,372 (British), viz.:

Every element found in Watson's structure appears in one or more of these references; and while Rendle's adaptation is limited to a "structure for horticultural purposes," and Coulson describes his as a "glazed structure," their pertinency, as already indicated, must be accepted if we regard Watson's efforts to have been exerted in the glazing art, rather than limited to such windows as are found on the sides of ordinary dwellings or buildings. To my mind, one seeking to perfect a structure wherein the mullion or the panes are easily removable would at once be referred to that branch of the art wherein that kind of structure was in greater demand, irrespective of the fireproof quality, and skylight and horticultural structures are in that branch. The contention of defendant, that the patent in suit is void for want of invention, must be upheld. It may be that Short and Sheldon, upon reproduction of their respective structures, are far from Watson in physical appearance; but, when the numerous later structures are added thereto, there is a sum total of mechanical, elements in the art which forbids Watson's claims upon any theory except that of aggregation or selection. Every feature is not only suggested, but worked out; and even if Watson's precise combination is not disclosed in any one alone of the many references, its production is, in my judgment, attained through the exercise of mechanical skill, and has not the quality of inventive novelty.

The third and fourth claims, in adding the element of a "metal" sash, must fall with the first two.

The conclusion is that the patent is void, and that a decree in favor of the defendant be entered accordingly.

W. Clyde Jones and Albert H. Graves, both of Chicago, Ill., for appellant.

John G. Elliott, of Chicago, Ill., for appellee.

Before BAKER, MACK, and ALSCHULER, Circuit Judges.

ALSCHULER, Circuit Judge. This is an appeal from a decree of the District Court holding void letters patent No. 702,754 to Watson, granted June 17, 1902, for certain "improvements in windows," and