CHANGE JN OCEAN LEVEL 207 



uplifted marine shore is 100 feet above tide. With this amount of visi- 

 ble uplift at the limit of latest giaciation, it would seem either that the 

 rise of ocean level had been small or that the uplifted area extends far 

 out under the sea and much beyond the loaded area. 



N'ova Scotia was wholly glaciated. Multitudes of huge erratics are 

 piled along the southeast shore and in the harbors (plate 17) ; but part 

 of the south shore exhibits no uplift (see the zero isobase of the ma])). 

 Either there has been no uplift here or else the flooding of the sea has 

 exceeded the uplift. 



On the coast of Maine the ancient beaches are from 300 to 400 feet 

 above tide. At Saint John, N'ew Brunswick, the uplift is 200 feet. On 

 the east coast of Newfoundland the apparent uplift is even greater, be- 

 ing at Saint John, according to Daly, about 575 feet. It appears prob- 

 able that the independent Newfoundland ice-cap spread widely beyond 

 the present shores, yet 575 feet seems a large amount of rise if some 

 non-visible rise is to be added. 



It should, however, be emphasized that no changes of sealevel, by wliat- 

 ever cause, can account for the great differential elevation of tlie slioro 

 features found in passing inland, rising from zero at Yarmouth and 

 Sydney to over 1,000 feet northwest of Quebec City. 



Tabulation and Desceiption of New Data 

 general discussion 



In the column of "definite" altitudes are placed only those measure- 

 ments which were taken on fairly clear features. The observations which 

 were uncertain, either for lack of time or lack of good datum or for poor 

 behavior of the aneroid, are placed in the other columns. The range 

 of error for the definite figures will usually lie within 5 or 10 feet. No 

 figures are placed in that column which are not regarded as correct within 

 5 feet. 



In many cases the altitude is quite precise and subject to no serious 

 change. The sources of error are: (1) wrong figures for the railroad 

 elevations, which were chief datum for most of the area in Canada and 

 for some territory in New England; (2) the variation of aneroid, which 

 was usually checked and corrected; (3) the uncertainty of the contours 

 of the older topographic sheets, where no precise altitudes are given 

 except the useless hilltops. A coincidence of errors is as likely to neutral- 

 ize as to magnify. 



The chief uncertainty in the study arises from failure, due to lack 

 of time or want of good altitude data, to determine closelv the initinl 



