994 THE GEOLOGY OF MINNESOTA. 



[Additional faote. 

 i 



consisted originally of a substance containing, say 50 per cent of silica and 25 per 

 cent of iron oxide, it has lost 25 per cent of its substance, and, without additional 

 supplies, it should show evidences of reduction of bulk to that amount. No such 

 contraction has been noted. On the contrary, Mr. Spurr discovered evidences of 

 expansile movements, owing to the introduction of foreign substances, chiefly siderite, 

 the whole increase in bulk, due to this cause, being estimated at "about one-tenth 

 of its former volume" (Bulletin x, page 163). 



A subsequent decarbonatizing he assumes to have been the cause of shrinkage 

 which produced the prevalent columnar jointing a phenomenon that we prefer to 

 attribute to shrinkage caused by a loss of heat. 



Again, in all cases observed microscopically, by Mr. Spurr or by the writer, 

 when siderite is associated with the oxide of iron (excepting limonite), it has been 

 found that the carbonate was later formed than the oxide. This assumed cause for 

 the present megascopic columnar structure requires, on the contrary, that the 

 carbonate precede the oxide. 



Additional facts. When the iron-bearing rocks collected about Gunflint lake 

 were first examined microscopically there were seen sundry things that suggested 

 igneous origin, and others which strongly indicated the agency of volcanic forces in 

 the vicinity. These were not fully understood at first, and were passed by as minor 

 anomalies which might be due to other causes. But as these signs multiplied and 

 were finally found to converge in an igneous rock, it was plain that it was necessary 

 to reopen the whole question of the origin of the iron-bearing rocks of the Mesabi 

 Iron range. It then became necessary to revise, before publication, the descriptions 

 of numerous rock sections, incorporating and interpreting the anomalous facts and 

 uniting them with other features observed later, and to point out more fully their 

 significance. Such revision and its results are included in the foregoing descriptions 

 (Part II), and they will be summarized briefly as follows: 



1. The rock No. 307 first attracted attention. It is described (page 309) as a 

 tuff, with a query, and is compared to a Carboniferous tuff of King's county, Ireland, 

 of which a small fragment had been obtained of Sir Arch. Geikie. This tuff is found 

 near the northern limit of the Animikie on a point on the south side of Gunflint 

 river, west of " the narrows." It is illustrated by the photograph, figure 9, plate V. 

 This indicated the existence of volcanoes in the region at the time of the opening of 

 the Animikie. This tuff is not silicified, and is believed to have rested on a land 

 surface. 



2. In rock No. 312, which is mostly composed of siderite, and which is repre- 

 sented from a photograph in plate VI, are found not only ragged pieces of rock like 

 No. 307, but also many pieces of jasperoid or gray flint, or devitrified glass (Nos. 



