206 



E. L. Troxell — Plaster of Paris. 



be any reason for an irregularity in that particular place. It 

 is significant, however, that both curves show the abrupt fall 

 with the gradual rise at the end, where they agree closely. 



Tests were made of the strengthening effect of gum arabic 

 when applied, to the surface of dry plaster in order to compare 

 with the original solution experiments. The results of the 

 surface application of the gum, although indicating a weaken- 

 ing effect in general, yet show that the hard coating may in 



Fig. 5. 



500 



- 400 



£ 300 

 o ZOO 



100 



Z 4 6 a 10 



Parts of Gum Arabic in 48 of water. 



/2 



Fig. 5. The average of the curves of fig. 4, which is based on experi- 

 ments with 108 briquets, soaked in solutions of gum arabic. 



the end add a little to the strength. The solutions of medium 

 density seemed to have the poorest effect on the briquets ; the 

 one-fourth solution averages nearly fifty pounds less strength 

 than the one forty-eighth, but it is slightly higher than the one 

 twenty-fourth solution. 



Five separate curves, based on the tested strength of 108 

 briquets, show a decided similarity of trend. Four of these 

 curves (A, B, C, and D, fig. 4), representing ninety-six briquets, 

 give results especially dependable. Each set of twenty-four 

 was made up at a single mixing to give an even texture and 

 quality to the plaster ; then they were numbered every sixth 

 to insure against local eccentricities. To make the experiment 



