PRESIDENT’S ADDRESS. 2 SRLS 
relationship, and :f they do, were the kindred even approxi- 
mately coeval? In the case of genera with varying and 
almost characterless species, such as natica and turrtella, 
how much reliance can be placed on the supposed identity of 
their fossil remains from wideiy separated localities! In 
considering che supposed widely-ranging forms, such as 
sharks and whales, cam we attach much weight to identifi- 
cations based on isolated teeth? Or, again, is the geciogical 
column so perfect in Europe thatour strata must of necessity 
fit into it somewhere? If we regard most or all of these ques- 
tions as settled, and it is doubtful if any of them are, then 
the conclusions based upon them do not always point in the 
same direction when we come te apply them to the deter- 
mination of the age of the Australian Tertiaries. 
The Lyellian method of determining the age of Tertiary 
strata may be simply stated. Find the percentage of recent 
mollusca in the beds. The application of this principle to 
Australian strata in anything like an exhaustive manner is 
due to Professor Ralph Tate. To a certain extent he used 
‘the comparative method as well, and, in fact, in his later 
writings this tendency, especially under the influence of 
Cossmann, becaine somewhat more pronounced. Still. it 
it was to the Lyellian principle that he always attached 
most weight. Since it is to Tate, above all others, that we 
owe our knowledge of the fauna of the marine Tertiaries of 
Southern Australia, it is but natural that his opinion 
should have. been generally accepted by recent workers in 
the field, and his influence in this respect was more potent, 
since the question most in evidence was not the cemparison 
of our beds with those elsewhere, but their connection one 
with another. Recently, Mr. Pritchard and myself have 
tried to divorce the wider question from the local one by 
the employment of local names for the formations (Hall and 
Pritchard, ’02.). Against the Lyellian method, the most 
constantly urged objection is that the personal equation 
enters so largely into the question of the determination of 
species that the results deduced are quite unreliable. Un- 
doubtedly there is a great deal of truth in this, and short as 
is the history of our detailed work among these beds, it is 
not without instances of very plainly-put differences of 
opinion on certain specific identities. But are we then of 
necessity to adopt the comparative method? There is 
certainly something very soothing about the term “com- 
parative method,”’ but in the present case is it really more - 
scientific than the Lyellian, and may not the self-same 
charge—neglect: to estimate che personal equation—be as 
