shew all the their relations by a linear series. Under these con- 

 ditions every system of rock classification must be more or less 

 artificial, and must be devised principally for convenience. And 

 the one that is on the whole the most convenient — that is the one 

 that eives the most information as to the composition and geo- 

 loj,'ical position of a rock, combinf^d with a nomenclature as little 

 cumbersome as possible- that system will })e tlie best. 



Tlio present paper includes the results of the microscopic 

 e.vamination of about L>r)0 rocks, but it is merely a commencement. 

 A rich harvest awaits petrologists in New Zealand, which will 

 take many years and many workers to gatlier in. Instead of 

 stating what is known or thought about the ago and geological 

 position of the rocks, which would have in\()l\ ed much controversy, 

 7 lia\H given full references to the reports, papers, and books in 

 which they are mentioned. All measurements are in inches or 

 decimal parts of an inch. Most of the specimens were collected 

 }>y myself, but a few ha\e been given me by friends, particularly 

 by the late Sir Julius von Haast! ^ 



Toa\oid adding to the existing confusion in nomenclature I 

 have followed Dr. A. Giekie and ^h. J. II. mis Te.ill in the larger 

 divisions of rocks, while in defining tlu- <'tnui» T lia\c taken Dr. 

 A. (hekie, xMr. Teal), Professor lionney arnl Pn-tes. ,r Wa.lsworth 



l.oiiM,,iH>iit!y T lias.' lu.t telt justifle 



