to dip and dig, the front edge of it wears off rapidly. The frame 

 furnished with the equipment was attaohed to the sample boxes as in- 

 tended by the manufacturer, but around this frame was placed a Formica 

 rectangle that received the pull from wires leading to a reciprocating 

 arm. This rectangle extended over the edges of the water pan upon 

 which it slid in slots cut in the Formica. Wet Formica has a very low- 

 coefficient of friction. Since the rectangle could not dip and since 

 it pushed the frame carrying the sample boxes at a point below the usual 

 center of rotation at which they dipped before the change in design was 

 made, the dipping was practically eliminated. Accordingly, the sponge 

 samples wore evenly. 



6. The samples were wired front and back to zinc diecast blocks, 

 which originally were the blocks bearing the bristles intended to scrub 

 paint samples* 



7» The amount of wet sponge extending below the edge of each box 

 varied with the softness of the sponge, in spite of the fact that all 

 sponges were cut wet to Ijr" x l|" x 3a" standard size. In a few cases, 

 extra soft sponges still allowed the box to hit the abrasive paper be- 

 fore the test was finished. In these cases, strips of plastic were in- 

 serted behind the sponge holder in the box. This insertion of plastic 

 was particularly necessary when a soft sample was being tested along 

 side of a stiff sample. 



8. Standardizing on sponge samples cut from the top surface so 

 that the samples were representative of the sponge but did not contain 

 any large holes. The samples were tested with the surfaoe against the 

 emery paper and usually were tested with one sample from one sponge 

 and with the other sample from another sponge, so that any large dif- 

 ferences between different sponges could be detected. The number of 

 strokes varied from !?00 to 1^00. The machine ran at 60 strokes per 

 minute. By means of weights in a pan attached to the top of the frame 

 fastened to the sample boxes, the pressure on the sample could be 

 varied. The initial pressure of 0.22 pounds per square inch was not 

 changed, since all the experimenter's time was spent in trying to get 

 more accurate data. The sponges travelled over a path of 10 inches 

 and were 3^" long, which left an actual rubbing path of 6| inches, or 

 a total travel over paper, in 1000 strokes, of 51|2 feet — or more than 

 a tenth of a mile. 



Procedure . — The data were obtained as follows: The wet sponge 

 was cut as described above, dried at Iii0o-1600F., weighed warm to 

 offset the rapid absorption of moisture from the air (a Rock Island 

 sponge, when exposed to 100 percent humidity, picked up Ul percent 

 moisture), rewet, wired to the block, abraded, and redried along with 

 any large pieces of sponge that may have been torn off during the test. 

 The wear was calculated to a standard 1000 strokes for comparison. 



U8 



