PROPOSED EXPIvORATlONS AND INVESTIGATIONS 277 



merary pair. In successive generations would they regress to the 

 mean, according to Galton, or diverge, according to Gulick? 



(b) Experiment similarly with two groups, not far from the 

 mean, of some very variable species, as with fish with anal fin rays 

 having a normal distribution from 8-14 with mode at 11, using the 

 one group with 10, the other with 12. Observe as above, 



4. Natural selection. 



The life-death value of variations. Subject large numbers of 

 Palaemonetes vulgaris (a marine species) to fresh water. Com- 

 pare the survivors with the dead, in respect to number of rostral 

 spines, because a fresh water species of the genus differs in having 

 fewer spines, as do the races of this species from brackish water of 

 very low salinity. 



5. Fecundal selection. 



(a) Grow a large series of several varieties of wax podded 

 beans, which self fertilize. Record fecundity. Is it correlated with 

 any other characteristics ? In the first series grow seed from the 

 five most fecund plants each year; in the second, grow seeds from 

 five plants of mean fecundity. Do the series diverge in other re- 

 spects than fecundity? 



(b) Do the number of visits of insects vary throughout the 

 flowering period of a species? If so, does the fecundity of the suc- 

 cessive flowers correspondingly vary? Sow from the most fecund 

 flowers and from those of average fecundity. Does the time of 

 flowering diverge in the two series ? Is any correlated character af- 

 fected ? 



(c) In a racemose species, do the lowest flowers difTer from 

 the highest in fecundity? If so, grow seed from each. After sev- 

 eral generations of such selection, do the plants differ? 



(d) Same with the inside and outside florets of a sun flower. 



6. Organic evolution. 



A tree frog which has the power of very slowly acquiring the 

 color of its environment may be grown in a green environment. 



Those which become the greenest would be artificially selected, 

 in the successive generations. 



In another lot, grown under ordinary conditions, select the 

 greenest in the successive generations. 



Is the green color acquired more quickly in the first case, 

 when the selection is along the line of individual accommodation, 

 than in the second case? 



