98 REPORTS ON INVESTIGATIONS AND PROJECTS. 



such biological relations for each species as capacity for self-fertil- 

 ization, either with or without artificial aid, adaptations for crossing, 

 number of viable seeds produced, etc. ; and, fourth, to find out 

 their capacity for successful manipulation, both as regards ease of 

 culture and of adequate control of the fertilization processes. 



Of the species that have been under observation this year, several 

 have shown notable variations, but the exact significance of these 

 can be known onh' when continued culture shall have shown their 

 behavior in inheritance. 



In the study of the laws governing fluctuating variations, material 

 has been collected for the investigation of secular variation in the ray- 

 flowers of Chrysanthe7ninn leucajithejnuvi, a small plot has been set 

 aside and fenced as a permanent chr^'santhemum preser\^e to supple- 

 ment the work I have done on Aster prenanthoides, and arrangements 

 have been made for continuing the studies on the latter species. 



The influence of self- and cross-fertilization and of the selection to 

 particular types upon fluctuating variabilit}^ and upon the fixation 

 of fluctuations is being studied in Zca mays, the character chosen 

 for this study being the number of rows on the ear. 



The relation of fluctuating characters of mutants to the corre- 

 sponding characters of the species from which they originated and 

 the question as to whether or not there is any tendency on the part 

 of the mutants to regress toward the parental condition has been 

 investigated in Oenothera lamarckiana,Oe. rubriyiervis , and Oe.nanella, 

 and the results were published as part of Paper No. 2 of the Station 

 for Experimental Evolution, entitled " Mutants and hybrids of the 

 Oenotheras," by D. T. MacDougal, assisted by A. M. Vail, G. H. 

 Shull, and J. K. Small. I have again collaborated with Dr. Mac- 

 Dougal and Miss Vail in a continuation of these studies, which will 

 be published soon in the series of papers from this station. 



Although experimental evolution is naturally divided into varia- 

 tion and inheritance, it must always consist in a combination of these 

 two processes, so that in most cases the investigations involve both 

 elements. Inheritance is always the test of mutations, and it is 

 likewise involved in every problem of selection ; but in the studies 

 already mentioned variation is the prime object, while in those that 

 follow, the laws of inheritance are the chief object in view. 



In 1904 an attempt was made to repeat the investigations of 

 Johannsen on the inheritance of seed-weights in beans, becau.se his 

 results were thought to be in disaccord with Gal ton's law of inher- 

 itance. The choice of Burpee's bush lima bean for the investigation 

 was unfortunate, because it does not maintain a strict bush form, and 



