Hcliaiifluis A mil (US Syiicotyleus. 473 



If all the seeds of a single head are planted out in 

 groups according to their degree of syncotyly, very little 

 ditterences will be seen in the results. Atavists and seed- 

 lings, in which the fusion extends less than half way up, 

 give a progeny the value of which is, on the average, 

 somewhat lower, but only if we fix our attention on the 

 mean values, and not if we compare the single individ- 

 uals. Many an atavist or hemi-syncotyl has a higher value 

 than most of the average syncotyls, and amongst these 

 latter, the question whether the fusion is so complete 

 that the apex of the double leaf exhibits no invagination, 

 has no effect on the hereditary values of its offspring, so 

 far as I have been able to determine. In the same way 

 the disturbances in the disposition of the leaves, which 

 so often follow on syncotyly, are of no value as selective 

 characters (§ 9) ; but it is not necessary to describe the 

 experiments which prove this point. 



Hemi-syncotylous seedlings are always rarer, and 

 often much rarer than the true syncotyls; amongst these 

 latter on the other hand, the highest degrees of symphysis 

 are more abundant than the lower ones, which have an 

 obvious invagination of the apex. I have often recorded 

 separately the various degrees of syncotyly in my seed- 

 lings. In this way figures are obtained which give curves 

 with two peaks, such as have been found for other anom- 

 alies, especially for fasciation.^ The atavists constitute 

 one peak and the most complete syncotyls the other. From 

 the former the curve drops rapidly to mount again grad- 

 ually with the increasing degrees of symphysis ; whilst 

 on the other side of the apex of the syncotyls there is 

 another rapid drop to the very rare cases of amphicotyly, 



V97/r Ics courhcs g,altomenncs dcs monsfruositcs. Bull. Scienti'f. 

 de la France et de la Belfrique. Publie par A. Giard. Vol. XXVII, 

 p. 396, April, 1896. See especially the curve on page 397. 



