368 



SCIENCE 



[N. S. Vol. LIII. No. 1372 



lels the different ooal beds of northern France 

 with those of Belgium and England. 



The greater part of the volume is taken up 

 with the insects (pp. 93-321), and the author 

 confirms Handlirsch's conclusion that during 

 Westphalian time hexapods were large, in fact, 

 that as a rule they were " giants." Pruvost 

 thinks that the Westphalian insects were not 

 all carnivorous, 'but that some may have fed 

 on the pollen, etc., of plants like the cordaites 

 and cycadophytes ; in other words, that the rise 

 of the insect world was largely conditioned by 

 the development of inflorescence among plants. 



Insect impressions, to be preserved in the 

 rocks, must be entombed in the very finest of 

 sediments. The author states that they are 

 found only in shales, in association with deli- 

 cate plant remains, and with those of animals 

 as well. The very best ones, of rare occurrence, 

 have, however, suffered no appreciable trans- 

 port or maceration, hut were buried quickly 

 along with the most fragile plants in the soft- 

 est of muds; while the majority of the speci- 

 mens found commonly in the " insect beds " 

 have undergone more or less long periods of 

 floating, and consequent maceration and disso- 

 ciation. The floated specimens occur at times 

 with stronger plant fragments and the remains 

 of animals, all in varying degrees of decompo- 

 sition. 



Pruvost breaks up Handlirsch's order Prot- 

 orthoptera, and puts the majority of his fam- 

 ilies in a new suborder, the Archiblattids (3 

 species deserrbed), which are present as early 

 as the base of the Westphalian. These are 

 '•■ the simplest and oldest of Protoblattoidea " 

 and they may have had their origin in the 

 Paleodictyoptera, the original source-stock of 

 all insects. Two other suborders of Proto- 

 blattids are erected, Mimoblattids (for Ameri- 

 can forms) and Archimantids (1 described). 

 The author remarlis on " the homogeneity and 

 antiquity of the blattid phylum," describing 

 43 forms, and on its early separation from the 

 rest of the orthopterids. Of Paleodictyopter- 

 ids he describes but 3 forms. He believes that 

 the greatest evolution of Paleozoic insects took 

 place during the Westphalian, and states that 

 at the top of the Lower Carboniferous (Dinan- 



tian or Mississippian) but one order is known ; 

 early in the Westphalian three orders are 

 " scarcely outlined " ; and at the end of the 

 Westphalian " almost all the Paleozoic phyla 

 are fully established." 



The evohition of insects was especially rapid at 

 the base of the Westphalian (Flines member), 

 again at the base of the upper part of the same 

 series (Ernestine), and at the top of the West- 

 phalian in the Edouard member. And this three- 

 fold acceleration in insect evolution is in harmony 

 with the floral enrichment. 



We must add here that the supposed insects 

 found in the Horton formation (early Missis- 

 sippian) of ISTew Brunswick, Canada, and 

 mentioned in the table opposite page 293, have 

 been shown to Professor H. F. Wickham and 

 Dr. David White, with the result that both 

 paleontologist and paledbotanist agree that 

 they are not insects but the caribonized frag- 

 ments of woody plants. 



To the young author, a favorite student of 

 Professor Barrois under whose direction are 

 being carried out a series of studies designed 

 to apply the " paleontologic method " to the 

 problems of the coal basin of northern France, 

 are extended our congratulations on his great 

 achievement. 



Charles Schuchert 



SPECIAL ARTICLES 



THE RELATIVITY SHIFT OF SPECTRUM LINES 



Three experimental tests of Einstein's 

 Relativity Theory of Gravitation have been 

 proposed. Two seem to have been verified 

 experimentally. The third, the predicted 

 shift of solar spectrum lines, is stiU very 

 much in dispute. Evershed and Eoyds,^ and 

 Schwarzschild- obtained very discordant re- 

 sults. St. John," with very fine apparatus, 

 also obtained very discordant results with 

 however a zero effect, on the average. Grebe 

 and Bachem* at first obtained discordant re- 

 sults, but a more careful analysis of their 



1 Bulletin 39, Kodaikanal Observatory. 



2 Sitzungshericlite, Berlin Akad., p. 1201, 1914. 



3 Astro. Jour., 46, 249, 1917. 



4 rerh. d. D. Phys. Ges., 21, 454, 1919. 



