390 II. L. Wells — Self -feeding Sjirengel Pump. 



That this metainorphism was of very early date, we are con- 

 fident because the syenite is found to underlie Tin conformably 

 the quartzites and slates of the Taconic or Huronian. The 

 change from non-crystalline to crystalline was probably pro- 

 duced at the time of the folding of the earth's crust which 

 turned both the Keewatin green schist and the Keewatin 

 syenite up on edge so that they stand vertical over a large 

 part of northern Minnesota. 



If the Saganaoa syenite be as supposed, the northeastern 

 extension of the Giant's Range, then this whole belt of rock 

 from the Mississippi river to the Canadian boundary and 

 beyond must be Keewatin and not Lauren tian. And indeed 

 there are indications from the relations of the Giant's Range 

 syenite to the vertical green schists on the south of the syenite 

 as seen near Hinsdale on the Duluth <fc Iron Range R. R, that 

 this is a correct inference. If this extensive range of gneiss: c 

 rock which has generally been supposed to be Laurentian is then 

 not truly Laurentian but Keewatin, it seems likely that in the 

 latter we have a rock come down to us unchanged through all 

 the ages, which in its various phases represents the original 

 fire crust of the earth more nearly than any other known rock. 

 But this is a question to be discussed at some future time. 



In closing. I wish to suggest an economic bearing of this 

 question. If the Saganaga syenite be of Keewatin age, and 

 contain chalcedonic silica in an original unchanged condition, 

 it is not unlikely' to contain also Keewatin iron ore deposits 

 free from titanium, and of high grade in other respects. It 

 can thus no longer be laid down as a law for explorers in the 

 northwest that the gneisses contain no iron ore deposits. It is 

 believed that the non-titaniferous magnetites of iNew York and 

 eastern Canada afford good illustrations of this class of iron 

 ores, and that they have been changed from chemically pre- 

 cipitated hematites of Keewatin age. 



December 20th. 1890. 



Art. XLIV. — On a Self-feeding Sprengel Pump; by EL L. 



" TTells. 



The arrangement here described for raising mercury 

 continuously to the top of a Sprengel pump, or for other 

 purposes, may be of use or of interest to some chemists 

 and physicists. The method, though exceedingly simple, seems 

 to be new. 



Fig. 1 shows the apparatus in its simplest form. Water 



