GENETICS: E. C. MacDOWELL 
291 
Co., 1887, (6). Knott, W. T., Rep. GeoL Marion Co., 1885, (11). Foerste, A., in Kindle, 
E. M., Chicago, J. Geol. Univ. Chic, 22, 1914, (709-711). 
2 Shannon, W. P., Indianapolis, Proc. Acad. Sci., 1895, (53-54). Moore, J., and Hole, 
Ibid., 1902, (216-220). Culbertson, G., Ibid., 1903, (202-205). 
3 Locke, J., Ohio Geol. Surv., 1838, (246). Orton, Rep. Geol. Surv. Ohio, 1, 1873, (377). 
Perry, N. W., Amer. Geol., Minneapolis, 4, 1889, (326-336). Foerste, A., Chicago, J. Geol. 
Univ. Chic, 3, 1895, (50-60, 169-197). Prosser, C. S., Ibid., 24, 1916, (456-475). 
* See Foerste, A., Kentucky Geol. Surv., Bull., No. 7, 1906. 
Dr. J. E. Hyde called my attention to this very interesting occurrence. 
6 Foerste, A. F., Amer. J. Sci., New Haven, (Ser. 4), 18, 1904, (321-342). 
^ See Foerste's Map of the distribution of the facies of the Clinton formation. 
' Cornish, V., London, Geog. J., 18, 1901, (170-202). 
* Hunt, A. R., On the formation of ripple marks. Proc. Roy. Soc, 34, 1882, p. 4. 
9 L. c, p. 190. 
^° Supan, A., Grundziige der Physischen Erdkunde, 1911, p. 260. 
Kjuemmel, O., Handbuch der Ozeanographie, 1911, vol. II, p. 383. See also 
Berghaus, A., Atlas der Hydro graphic, 1891, PI. XX. 
*2To what extent inorganic or organic precipitation may have played a role I am in no 
position to judge. Observation 7 seems to point to it as a source of finely divided cal- 
ciumcarbonate. 
"Kriimmel, O., loc. cit., p. 112. 
THE BEARING OF SELECTION EXPERIMENTS WITH DROSOPHILA 
UPON THE FREQUENCY OF GERMINAL CHANGES 
By Edwin Carleton MacDowell 
STATION FOR EXPERIMENTAL EVOLUTION. CARNEGIE INSTITUTION OF WASHINGTON 
Communicated by C. B. Davenport, March 5, 1917 
Biologists generally agree that changes in the germ plasm do occur. 
On the other hand, there appears to be considerable disagreement re- 
garding the frequency of these spontaneous changes. Some experi- 
ments, as illustrated by the work of Pearl, Hj. Nilsson, De Vries, Tower, 
and Johannsen,^ are more easily analyzed by supposing that changes in 
the germ plasm occur very rarely in comparison with the number of 
generations of individuals studied, so that the origination of new races 
by selection is not generally possible. Other experiments, notably those 
of Castle, Smith, Middleton, and Jennings (on Difflugia) ,^ have been in- 
terpreted as showing that such changes are occurring so frequently that 
they may be found in each generation, and so afford a basis for selection 
to make continuous progress. As long as different experiments lead 
their authors to such different conclusions, no broad generahzations as to 
the scope of the evolutionary significance of selection may be drawn 
without the most intimate and critical consideration of all other related 
investigations, and, accordingly, all additional evidence that may be se- 
cured has an important bearing. 
The familiar Mendelian units are currently conceded to arise sud- 
