Instead of going home as directed, I for hours had 

 my place in the empty bucket line. While this was 

 going on Garrett was having section after section of 

 hose attached to both the Philadelphia and the 

 Diligent engines, only to burst before the water 

 reached the upper floors. Fortunately the Resolu- 

 tion Company had a number of new sections of hose 

 which were pressed into the service, and were at- 

 tached next the engines where the pressure would 

 be greatest. The floors were flooded, the attics and 

 roofs fell into the water, and spreading fire was 

 stayed. The success of this expedient was so com- 

 plete that it turned the attention of all firemen in 

 the direction of securing more reliable hose. . . .[i] " 



At the time I am writing of the old firm of N. & 

 D. Sellers, that dated back to the Revolutionary War, 

 was carried on under the same name by my father, 

 son of Nathan, and his two cousins Samuel and James, 

 sons of David. James was an active member of the 

 Philadelphia Hose Company, and was much inter- 

 ested in experimenting to find a substitute for the 

 hemp or flax thread for sewing hose, not subject to 

 their decay, and that would not be injured by the 

 dubbing and have a lifetime at least equal to that of 

 the leather. 



He made a short section of hose sewed with an- 

 nealed copper wire with a double row of stitching. 

 To have the wire sufficiently pliable to draw the 

 leather close and make a tight seam he was obliged 

 to use a wire considerably thinner than the hemp or 

 flax thread, and owing to the difficulty in inserting 

 the wire in the punctured holes they were made 

 larger, and the leather did not close tightly on the 

 wire, which lacked the swelling property of the thread 

 stitches which had aided greatly in closing the holes. 

 This section was tried fully up to the required pres- 

 sure, but not without considerable steeping through 

 the stitch holes; at an increased pressure the thin wire 

 cut the leather and the leakage increased. Yet the 

 experiment was considered so far a success that the 

 hose company agreed to bear the expense of further 

 experiments, and Mr. J. Morris and Abraham L. 

 Pennock were appointed a committee with James 

 Sellers to assist him in prosecuting the experiments. 

 Many were tried to render the wire-sewed seams 



25 Bracketed numbers refer to sources of text material for the 

 preceding section. List of sources is given on p. 193. 



more pliable and to stop the cutting and steepage, 

 but unsuccessfully. 



The next experiment was by nailing: this at first 

 promised well. A workman, named Andrew Link- 

 felter, was in the employ of the firm [N. & D. Sellers], 

 whose business for many years had been driving tacks 

 in putting together hoops or rims of meal and grain 

 sieves, and who prided himself on being able to so 

 drive a large headed wrought tack, then called a 

 "clout nail," so as to insure a certain and perfect 

 clinch always in the same direction. To him was 

 given the job of putting together the first sections of 

 nailed hose. This was done by lapping one edge of 

 the leather over the other as it is now done with the 

 riveted hose, and driving two rows of clout tacks, 

 clinching them on an iron mandrel. Several sections 

 of 50 feet each of hose were so made, and subjected 

 to very severe tests, showing no signs of yielding, but 

 in after tests the clinches began to give way, and 

 finally the seams ripped. This was mainly owing to 

 the swelling of the leather when wet and shrinkage on 

 drying tending to strain and loosen the clinch. 



The next experiment was driving one row of tacks 

 near the outer edge of the leather as it lapped over, 

 clinching them on the iron mandrel as in the previous 

 case; then a row of holes was pierced near the edge of 

 the inner lap and the tack inserted from the inside 

 with its head resting on the mandrel. The leather 

 was then driven down around the tack by a hollow 

 punch; a row of about a foot in length of these tacks 

 being so arranged with their points exposed, they 

 were turned by a pair of round-nose pliers, clinched 

 and driven home by light hand hammer. This stood 

 the tests well, but it was a tedious and costly process. 



My recollection is that during all these experiments 

 James Sellers devoted his whole time and attention 

 to them. Mr. Pennock was almost daily at the shop 

 for a short time, and Mr. Morris only occasionally. 

 I recollect one day when all were present. My father 

 asked if it would not be possible and better to put 

 another head on the tack instead of clinching it. 

 The next move that I remember was seeing my cousin 

 James casting a hard pewter composed of tin, lead 

 and copper into a single rivet or washer at a time in 

 molds similar to the single mold used for casting rifle 

 bullets. In this way the rivets and washers were made 

 for the first section of riveted hose. 



The next advance was in making long brass molds 

 in which 1 5 or 20 rivets or washers in a row were cast 

 at a single pouring. They were afterwards cut from 



8 



