:-^78 Transactions.— Geology. 



to interest persons who are not professional geologists — I wish 

 to make a few remarks on the nature of the evidence by 

 which geologists correlate different strata and ascertain their 

 relative ages. 



This evidence is of three kinds — (1) stratigraphical, (2) pa- 

 Iccontological, and (3) lithological. Stratigraphical evidence, 

 when complete — that is, when one stratum is seen actually to 

 overlie another without any possibility of this being due to 

 inversion or overthrust — is absolutely decisive as to the rela- 

 tive ages of two sedimentar}'- rocks. Unfortunately, complete 

 evidence can rarely be got. Usually the geologist has to infer 

 the relative positions of two rocks by putting together two or 

 more disconnected sections. This lets in the possibility of an 

 unobserved discordance — that is, a fault or an unconformity — 

 occurring between the sections, and so reduces immensely the 

 weight of evidence. Stratigraphical evidence, to be conclu- 

 sive, usually requires that a considerable area of country 

 should be mapped in great detail ; and this, again, requires a 

 more minutely-accurate topographical map than exists for any 

 ])art of New Zealand. Until we have geological maps equal 

 to those of the British Islands we must generally receive with 

 caution, and treat as probable, the results arrived at by strati- 

 graphical evidence alone. 



With pakeontological evidence it is quite different, for bere 

 good maps are not necessary, and the fact of two localities 

 being disconnected is of little importance, provided they are 

 both within the same biological province. But palaeontology 

 bas its difficulties. First, we have the danger of mixing to- 

 gether fossils from different horizons although collected close 

 together ; secondly, mistakes may occur from labels having 

 got displaced or mixed, or perhaps forgotten altogether. In 

 tliese cases the attempt to identify the localities by memory 

 has often led to mistakes. Thirdly, there is the difficult}- of 

 correctly identifying species. All these are real difficulties 

 which may prevent accurate results being attained. Still, it 

 has turned out over and over again that when stratigraph}' 

 and palaeontology have been at variance the mistake has been 

 in the stratigraphy ; and this is a clear proof that, on the whole, 

 })ala5ontological evidence is the more reliable. After long and 

 detailed study of tbe geology of l-^urope and North America 

 by many geologists, tbe -only case where stratigraphy and 

 paleontology are still opposed is the so-called Colonies of 

 Barrande, in Bohemia ; and even here it is thought by many 

 geologists that the stratigraphy is in error. 



Lithological evidence is of two kinds — ^(1) included frag- 

 ments, and (2) mineral composition. Included fragments of 

 another rock are invaluable for proving relative age when the 

 fragments can be recognized with certainty. Mineral compo- 



