J^h'isfDoenr ('han(/<-^ <{f Level. ^/)i'(r <-<■>• , 37 
when the submcrgenct' formed a njvrrow strait across the Chig- 
aiecto isthmus the figures for the localities at St. Jolm an<l in 
Nova Scotia are perhaps some ."> feet too high. Tf, :is is prob- 
able, the levelings based upon raihvtiy altitudes refer to m<^an 
•tide, the}' also must l)e lowered nbout "> feet. 
The levelings were all ninde with Swedish hiindlevels con- 
structed l)y Wrede and Klfving, which contain a mirnjr held ver- 
tically by an ndjustnl>l(> weiglit ;ind sheltered from the wind by a 
little wooden case. l>y iiid of a scnle angles can :dso l»e me:is- 
^ired, and T have often mnde good use of them as a clii'ck and 
ixlso for plane table work. 
The liarometrical measurements were made with two aneroids 
from Nandet in l*aris. and each is Itased upon a series of 10 to 
25 observations l>y means of which the changes in pressure dar- 
ing the day ai'e grai)hically constructed, and from the ditferences 
thus obtained the hight is reckoned with due corrections for tem- 
perature. Though T have often in this way got remarkably good 
results, it is very desiral)lc that thesi' nieasui'cments should be 
checked by the spirit level, as figures should not be considered 
conclusive which have not an accuracy within thret' feet or about 
•one meter. 
('oNci.rsioNs. 
All the observations on the hight of the marine limit have 
been put down on a general map." and with the aid of interpola- 
tion isobases have been drawn through e(|iudly u|)lifted points 
with an interval of 2(M» tVct (('.(I m.). 
Concerning the extension of the isobases into the interior of 
the continent, where the nmrine limit could not be dii'cctly de- 
termined. 1 have tried to use interpolation in the following man- 
ner. As has l)een stated, it is proli;ible that the geoid-surfaee, 
which in the submerged regions is marked l)y the mai'ine limit, 
is situated in the trac^t noi'theast of lake Ontario at a hight some- 
what less than 7.") per cent, of the older high-water level recorded 
by the Irociuois l»eaeli. i-'iom tlie ligui'es given liv professor 
Spencert we find that this lieaeli is situate<l at about ."!(! per cent, 
of the llidiicwav bi'aeli at the thi'ee localities where liotli (teeur 
■"•Presented Willi tiiis i)iii)er in tiie I'roeeeilings of the I'.ostori Soci- 
ety of Natural History, but not here re|)nK[ucc(l. 
t'" \\\i^\\ level shores in the region of the (ireal b:ike> ;iihI tiieir ch-- 
ionnntion." .\ni. .b)ur. Sci., March. !S!»1. 
