Twenty- SECOND Annual Meeting. 45 



gards alternating currents, it seems to us highly probable that after some years the 

 wires will show unexpected changes in brittleness and tensile strength. We know 

 scarcely anything yet about heavy alternating currents. 



Certain facts, however, in regard to the electrical qualities of an insulating sheath 

 may be determined by tests. With this object in view, experiments were made in 

 the spring of this year, in the physical laboratory of the State University, upon six 

 of the principal arc-light wires in the market. All of them had been in outside use 

 by the Lawrence Electric Light Company. These tests were made to determine the 

 insulating power of the wires against leakage when exposed to moisture. The wires 

 were No. 6, American wire gauge, guaranteed by the makers to be water-proof. 

 Lengths of about 25 feet were taken, wound into coils, and immersed in large 

 tubs filled with hydrant-water. The two ends of each coil projected about two feet 

 out of the water. The coils were left in the water for nearly three months, measure- 

 ments at first being made every day so long as considered necessary. The mode of 

 testing was that known as the substitution method. One of the projecting ends of 

 a coil of wire was connected to a delicate Wiedemann adjustable-coil dead-beat mir- 

 ror galvanometer and to the positive pole of a single gravity cell. The negative 

 pole was connected by a wire to a copper plate 20 inches square, which was placed 

 in each tub of water, in the center of the coil. The other projecting end of each 

 coil was left insulated in the air. Resistance boxes reading to 100,000 and 33,000 

 ohms, respectively, were so connected that they could be substituted in the circuit 

 of the galvanometer and battery in place of any particular coil. For convenience, 

 a special mercury switch was so constructed that the observer could throw in suc- 

 cession the different coils or the resistance boxes into the galvanometer circuit, and 

 by equal deflections the resistances of the insulation sheaths could be determined. 

 It is evident by the arrangements of the experiment, that only through the insulat- 

 ing layer of the wire could the current escape to complete the galvanometer circuit; 

 and from the reputation of the wires, only a small leakage was to be expected. The 

 insulation resistance should be several million ohms per mile. We have embodied 

 our results in the form of curves indicated in the following diagram. (Explain 

 curves.) After three months' immersion final tests were made, which gave the fol- 

 lowing results: 



No. 1 7,200 ohms per yard. 



No. 2 8,067 ohms per yard. 



No. 3 816,000 ohms per yard. 



No. 4 7,333 ohms per yard. 



No. 5 3,702 ohms per yard. 



No. 6 3,300 ohms per yard. 



As the result of these tests, No. 3 is now used almost exclusively on the Lawrence 

 electric light circuits; and at this writing, after nearly a year of oat-door exposure, 

 the insulation seems as perfect as at first. 



ON BARITE AND ASSOCIATED MINERALS IN THE CONCRETIONARY 

 ROCKS OF EASTERN KANSAS. 



BY PROF. E. H. S. BAILEY, AND E. E. SL0S30S, LAWRENCE. 



[Abstract.'] 



These rocks are found as boulders upon the hills in the vicinity of Lawrence, 



especially upon Mount Oread. It seems probable that there is a layer of these rocks 



below the surface. Similar boulders are found in Jefferson county along the line of 



the narrow-gauge road, in some places beneath layers of clay and above the lime- 



