583.111 179 



III 



Second Contribution on Numerical Variation of parts 

 in Ranunculus repens (L.) 



THE present investigation was undertaken in continuation of a 

 paper which appeared in Natural Science for May 1897, p. 

 323, as it was thought that the conclusions drawn and results 

 obtained from a series of 500 specimens, might be influenced by 

 the small number examined. Accordingly (as noted in Natural 

 Science, June 1897, p. 429), another series of .500 was examined, 

 and in the following paper the results of this second series, 

 separately, and combined with the first, are discussed. It will, I 

 think, be admitted that conclusions based on observations of 1000 

 specimens are fairly well founded. The specimens were again 

 most kindly supplied by Miss K. M. Hall, Curator of the 

 Whitechapel Museum, from practically the same locality as 

 last year, my occupation preventing me from collecting them 

 myself. 



The number of specimens examined was again 500. The 

 method employed in recording the results detained was slightly 

 different from that used last year. Sheets of paper ruled in five 

 columns were taken, and the columns were marked successively — 

 first column, consecutive numbers from 1 to 500 ; second column, 

 <5alyx ; third, fourth, and fifth, corolla, stamens, and carpels respec- 

 tively, each specimen having one line to itself on one sheet. 



Example — 

 1896. 



Number. Calyx. Corolla. Stamens. Carpels. 



15 5 68 46 



2 6 9 55 43 



With this method the results are more accessible than with the 

 former plan, which, however, possessed certain advantages. 



The total number of parts in the 1000 specimens examined 

 was — 



