198 H. H. NEWMAN. 



known to be a process involving a precocious dichotomous bud- 

 ding initiated by the primitive ectoderm. The individual fetuses 

 are produced agamically as bud products and therefore constitute 

 a clone; so whatever somatic variations occur among the four 

 fetuses of a set are obviously types of clonal variation. 



Bud or clonal variations have for a long time been propagated 

 vegetatively by cutting, grafts, etc., and are known to hold true 

 to their somatic characters. The explanation of bud or clonal 

 variations that has usually been offered involves the introduction 

 of two causal factors, one internal and the other external. It is 

 said that a local peculiarity may be induced by local conditions 

 outside of the bud tissue itself, such as unfavorable or extra 

 favorable position on the plant; but such an explanation involves 

 many theoretical difficulties. A more acceptable explanation 

 involves the assumption that there is some segregative mechanism 

 that brings about an uneven distribution of inherited factors, so 

 that one type of factor may be present in one part of the growing 

 plant and absent in another. Special cases of somatic segregation 

 are those cases that we have been accustomed to consider under 

 the term particulate or mosaic inheritance. Such inheritance 

 conditions are supposed to be due to a segregation in a single 

 hybrid offspring of the opposed characters of the two parents. 

 When, for example, a cross is made between a plant with white 

 and one with red petals, we may get a hybrid possessing flowers 

 with some white and some red petals. This is said to be a result 

 of the segregation of parental color factors so that some cells 

 get certain factors and others do not. 



So far as I am aware it has not been shown whether in bud or 

 clonal variations, germinal segregation goes hand in hand with 

 somatic segregation. If a green plant that produces a white bud 

 variation should be found to produce flowers on this bud that 

 bred true to the white character, we would have a very clear 

 case of parallel somatic and germinal variation. 



Now in the armadillo we have precisely this state of affairs, 

 as can be brought out by a study of the inheritance of double 

 bands and double scutes. For, when some of the fetuses of a 

 set exhibit the anomaly as shown in the mother and others do 

 not, we have unequivocal somatic segregation; but we have 



