326 Knopf and Sehaller — Two New Boron Minerals. 



persistent green color to the flame when heated with potassium 

 bisulphate and fluorite. 



The samples of lmlsite first submitted for chemical inves- 

 tigation were apparently homogeneous. They were, however, 

 strongly magnetic, and though it was suspected that they con- 

 sisted of a mixture of a non-magnetic borate with magnetite, 

 various tests that were made failed to show the presence of 

 the latter mineral. The very finest powder suspended in 

 water was completely attracted by a magnet and numerous 

 chemical tests failed to lead to a separation of the magnetite 

 from the borate. After the analytical work on these samples 

 had been finished, some specimens of hulsite, collected in pre- 



Fig. 1. Hulsite (opaque mineral) intergrown with vesuvianite and caleite. 

 Magnified 12 diameters. 



vious years, were found, which when tested were non-mag- 

 netic. These specimens showed that the earlier ones contained 

 admixed magnetite, and a careful examination of all of the 

 samples on hand showed the very frequent presence of mag- 

 netite in varying amounts included in the hulsite. Analyses 

 of the non-magnetic and magnetite free hulsite serve to estab- 

 lish its formula, and on the basis of these analyses the data 

 obtained on the first samples become intelligible, the samples 

 analyzed being about 1/5 magnetite. "While there were a 

 number of specimens of hulsite which were not magnetic and 

 therefore free from magnetite, several of them had become 

 more or less altered to a limonitic substance. Only half a 

 gram of pure material, unaltered and non-magnetic, could be 

 obtained. This was divided into three portions of about 1/6 

 gram each for the following analyses, the size of the grains 

 being <60> 100 mesh: 



