14(J Cooperative Inrcstifjatiotis on Plantn 



in situ, in a small picce of tissuc papcr. The buds dricd and wcrc prcscrved for 

 wceks, to bc countod at Icisure. Even if a petal ur scpal feil oflf it was preservcd 

 in thc papcr wrapper. \Ve believe it is largoly due to this nicthod of gathering 

 that our results show such a totally dift'eient distribution of scpals and petals to 

 thoso of other observers. 



Take thc casc of scpals. In thc Gucinscy scrics \vc found ouly thrce in- 

 dividuals with Icss than thrce scpals. On cxauiination linder the niicroscope 

 in one of thesc cases an abortive scpal was found, in another the sepal had 

 clearly oncc been attachcd, but in thc third Civse thc hcad was unfortunately lost 

 before microscopic cxauiination. Wc think Ihat wc niay .safcly affirni that no true 

 case of less than thrce scpals was fouml in tiic Guernsey plants. In the Dorset- 

 shirc gathering no cases of less thaii thrce scpals occurrcd. In thc Surrcy 

 gathering thcre werc in the matcrial six cases of less tiian thrce scpals, and 

 in five out of these six cases thc rulc of gathering bmis duly had been disregarded 

 by a young collector* ; and thc sixth was an abortcd Hower in which thc staniens 

 wcrc not propcrly developcd and thcre werc no pistils at all ! In the Bordighcra 

 celandiues tiicre arc only two cases of less than three scpals. These have not 

 been excluded from the calculations, becausc we had not whcn counting them 

 leamt from thc Doreet and Gucrnsey scries to regard all cases of less than thrce 

 sepals with grave su.spicion and cxaniinc such cases undcr the microscope. But if 

 in 624 cases only two such individuais occurrcd, it seems cxtremely probable thivt 

 evcn the.sc were cases of scpals knockcd off or abortcd. In a total of 21 -iO hcads, 

 thcre were 11 cases of less than thruc scpals, two of these wcrc abortcd ilowcrs, 

 fivc of them wcrc old Huwcrs witii parts loosc, one had ouce had a sepal which had 

 been lost, and thrce were not closely examincd. The experiencc we have had 

 leads US to believe that cach flowcr ought to be gathercd as a bud and at once 

 wrapped up. It seems to us that all we cau admit is a possibiliti/ of three heads 

 with less than three sepals in 214!) cases, while Dr Ludwig's 3000 from Greiz 

 show no less than 60 defiuitc casesf. We feel fairly confidcnt that had the 

 Greiz flowers been collccted as buds and possibly gathercd and counted by adults 

 instead of school children they would not have differed so widely from our material 

 in this respect. As far as our experiencc rcachcs, wc ((uestion the existence 

 of any normal flower with less than thrce sepals. This vicw may be modified 

 whcn fiutiii'i- matcrial from central Europc gathcnil in bud and, it' necessary, 

 microscopically examincd, is availablc. Wc hope next se;vson that this problem of 

 the sepal may bc directly iuvcstigatcd. 



(2) Material of the present iiivesti<jation. 



(A) 624 heads gathercd in a vincyard at Bordighera, Italy, between March 4th 

 and 7th. Due to Mr Francis Galton. The sepals, petals, staniens and pistils 

 wcrc counted by .Mr N. Blanchard. Thc tables were prepared and the Statistical 



" The caloulntor's notes nre : No. 51, "only one sepal und tliis hnriKing loose" ; No. 52, " old flower 

 with petals detachcd " ; No. 53, "riatil» looso"; No. Ol, " very old llower, all parts loose"; No. 69, 

 "vcry old flowor." Xo remarks as to age of this Icind occur in other cases. 



t See Biomelrikii, Vol. i. pp. 13 — 1.5. 



