64 cross: lavas of Hawaii 



of Hesse, or to dolerites, or to "diorites" of New Zealand and Aus- 

 tralia; the essexitic andesites of Haleakala find their nearest 

 relatives in "mugearite" (Harker) from the island of Skye, in 

 " trachydolerite" from Cripple Creek, Colorado, in "carmeloite" 

 (Lawson) of California, and among the rhombic prophyries of 

 Norway. 



One of the lavas of Kilauea is very close to a "diabase" from 

 Cape Colony, Africa; another is most like a "kersantite" of Thu- 

 ringia ; a third is nearest to a basalt of Cerro San Miguel, Mexico ; 

 a fourth may best be compared to a basalt of Pantellaria or one 

 from Cockburn Island. Space is lacking to present further details 

 in this place, but what has been said illustrates the widespread 

 distribution of the magmatic types which are nearest to the lavas 

 of Hawaii in chemical composition. 



With respect to conditions of occurrence it is plain that the 

 closest analogues of the Hawaiian rocks occur partly in areas of 

 folded mountain chains, partly in regions of faulting with sub- 

 sidence, partly in districts where either such characterization 

 of occurrence is little justified. Whatever be intended by the 

 terms Atlantic and Pacific branches or "Sippen" the Hawaiian 

 rocks do not fall clearly into line with either. Among them are 

 some typical "alkali" or "Atlantic" types, some unquestionable 

 "subalkali" or "Pacific" varieties, and a number of intermediate 

 forms. 



The Hawaiian magmas tend to show that the generalizations 

 as to geographic distribution or the genetic relations of the 

 "alkali" and "subalkali" groups included in current definitions 

 of the "Atlantic and Pacific branches" or "Sippen" are far from 

 correct. Hence in their present forms they can have no place 

 in a petrographic system. 



PALEONTOLOGY. — Remarks on the fossil turtles accredited to 

 the Judith River formation. F. H. Knowlton, Geological 

 Survey. 



The similarity, or often apparent identity, between the turtle 

 faunas of the Judith River and Lance formations has been em- 

 ployed of late, and by several paleontologists, to prove the Cre- 



