28 G. J. Brvsh—Gahnitefrom Mine Hill, N. J. 



erlv stars proportionally more), and l"-2 in declination. This 

 indeed is completely sufficient. It would follow from this 

 that the probable error of a mean of two observations is 0^-04 

 and 0"-05. But with the above mentioned mean diiferences 

 among 22 chances the differences 0»-25 and 3" would occur 

 once, and among 142 chances those of 0s"34 and 4''1 once. Ac- 

 cordingly it would be advisable perhaps as often as differences 

 occur which exceed O^'S and 3 "-5 to make a third observation 

 for verification. 



The observer will do well to inform himself completely on 

 the subject of the introduction of the D. M., found in Bonn Beob., 

 III. It is there noticed that sometimes two very close faint 

 stars by their united light have given cause to the introduction 

 of a 9=" star which does not exist ; such are to be observed if 

 bright enough. If a 9"' star should be found which is not in 

 the DM., and these cases are very rare, it is to be observed and 

 the fact carefully noted, to be published in the final catalogue. 



To secure the internal connection and completion of the 

 whole, the participants are urgently requested to transmit, be- 

 fore July of each year, annual rejjorts on the work of the year 

 to the President, or one of the Secretaries, of the Society. 



Art. VI.— On Gakaite from Mine Hill, FranUin Furnace, New 

 Jersey ; by Geo. J. BRUSH. 



The rare species Gahnite has been again found at a new 

 locality in a cross-cut made by the New Jersey Zinc Co. from 

 the valley of the Wallkill river to an opening on the south 

 end of Mine Hill. I collected specimens at this locality in the 

 summer of L869, and by blowpipe examination at that time 

 determined the mineral to be a zinc spinel. 



The mineral differs in its crystalline characters from the speci- 

 mens of other localities in the frequent occurrence of the cubic 

 plane ; in fact the cubic planes are often the largest, so that 

 the crystals are cubes with truncated dodecahedral edges and 

 only small octahedral planes. There are also minute planes of 

 the trapezohedron 2-2, truncating the edges of the clodecahe- 

 dron; also others of the trigonal-trisoctahedron 3. Besides 

 these there are sometimes two planes between th(> cubic and 

 the octahedral, which appear, from examination ;in<l appi'O^i' 

 mate measurements hj Prof. Dana, to belong to the Ibnu.s ^-^ 

 and 8-8. Their surfaces are rounded, and feeble in luster, and 

 generally they are blended in a single curved plane, onseiiuentlV 

 the measurements are not entirely satisfactory. The inclinations 

 on a cubic plane, obtained by Prof Dana, aVe for 4-4 160° 30', 

 for 8-8, 170° 30'. 



