NOVEMBEE 22, 1895.] 



SCIENCE. 



699 



physiological deductions. The experiments 

 brought out three important facts, namely, (1) 

 adaptation (20-30 minutes stay in darkness) 

 aiFects the sensibility for colors unequally. Be- 

 ginning at zero for the red, the improvement 

 increases as the wave-length shortens till for 

 the violet it is very considerable ; (2) adapta- 

 tion does not make the colors seem more intense 

 as colors, but only more luminous, as if white 

 light had been added ; and this may reach such a 

 pitch with very faint lights that the colors are 

 wholly lost in the white light ; (3) the sensi- 

 bility of the /oyea is unaffected by adaptation. 



On these facts Pariuaud bases a theory of 

 the rods and cones and the visual purple. In 

 the fovea there are cones only, and, as every- 

 where, they are without purple. Adaptation 

 appears to be an affair of the rods and the pur- 

 ple ; it takes place where they are found, and 

 fails where they are absent. Since the lumin- 

 osity alone is affected, it is natural to regard 

 them as an end-organ for luminosity only, leav- 

 ing the cones to mediate color. The matter is 

 not so simple, however, as a mere separation of 

 the organs, for the cones must also mediate 

 white, and, indeed, in Parinaud's opinion, could 

 do nothing more than that without the coopera- 

 tion of the cerebral centers. Hemeralopia 

 (night-blindness), which appears to be due to 

 a deficiency in the purple, confirms this theory 

 of its function, as also does the good develop- 

 ment of the rods and purple in the eyes of noc- 

 turnal animals. The purple is able to increase 

 the effect of faint lights because of a fluorescent 

 or phosphorescent property. Parinaud's argu- 

 ments for such a property make a very plausi- 

 ble case. If he is correct the purple becomes 

 an agent for the actual production of light on 

 faint luminous stimulation instead of an agent 

 for increasing the irritability of the visual ap- 

 paratus. The paper concludes with a fairly full 

 account of the work of other observers in related 

 lines.* The contribution is important in bring- 

 ing together a number of more or less disre- 

 garded facts and showing their very great physi- 

 ological significance. 



E. C. Sanford. 



*The reviewer hastens to withdraw his criticism of 

 the first part of M. Parinaud's paper for deficiency in 

 this respect. 



Geology of the Green Mountains in Massachusetts, 

 By Kaphael Pumpelly, J. E. Wolff and 

 T. Nelson Dale. Monograph XXIII of the 

 United States Geological Survey. 1894. 4°. 

 Pp. xiv, 206. Plates 23. Price $1.30. 

 The monograph before us is the most detailed 

 and valuable contribution yet made to the solu- 

 tion of the much debated ' Taconic question,' 

 than which none other has achieved greater 

 prominence or excited more bitter feeling in the 

 last fifty years of American geology. 



Since the discovery of actual fossils in the 

 metamorphosed strata of Vermont by the Rev. 

 Augustus Wing, the labors of many have indi- 

 cated the true relations that are now demon- 

 strated, yet nevertheless the difficulties of the 

 problem were so great, and the tendency to 

 generalize without detailed field work had been 

 so marked, that Mr. Pumpelly and his co-labor- 

 ers decided to throw aside all previous conclu- 

 sions and by detailed and patient observation, 

 based upon topographic maps in a crucial area, 

 to trace out step by step the relations of these 

 much disturbed and metamorphosed sediments. 

 Accordingly the northwest corner of Massachu- 

 setts was selected and study was focused especi- 

 ally upon Hoosac Mountain on the east. Grey- 

 lock Mountain on the west and the valley bet- 

 ween. Hoosac Mountain, well known for the 

 famous tunnel that penetrates it, is an anticlin- 

 orium with a core of granitic pre-Cambrian 

 gneiss (the Stamford gneiss), on which rests, with 

 conformable lamination, another variable white 

 gneiss that is at times a recognizable conglom- 

 erated and even a quartzite (the Vermont forma- 

 tion). Above the last and still conformable is 

 a great thickness of albite schist (the Hoosac 

 schist), which is itself succeeded on the east by 

 the Rowe schist. The Vermont formation is 

 Cambrian ; the Hoosac schist is Cambrian below, 

 Silurian above. The Rowe schist is Silurian 

 and of minor importance in the problem. On 

 the west side of Hoosac Mountain the Hoosac 

 schist fails and the Vermont formation runs 

 under the Cambro-Silurian Stockbridge lime- 

 stone that has been degraded to form the valley. 

 It should be remarked that all the strata of 

 Hoosac Mountain proper, except the Stamford 

 gneiss, are metamorphosed elastics. 



Greylock Mountain, with its spurs, is a 



