Systematic Atavism 639 



trum, being limited to the base of the spike, 



I decreasing in size from the lower flowers 

 yard. Connected with these atavistic bracts 

 a feature of minor importance, which how- 

 'T, by its almost universal accompaniment of 



bracts, deserves our attention, as it is indica- 

 B of another latent character. As a rule, the 

 lets are grown together with their axillary 

 s^er-stalk. This cohesion is not complete, nor 

 it always developed in the same degree, 

 metimes it extends over a large part of the 

 organs, leaving only their tips free, but on 

 ler occasions it is limited to a small part of 

 > base. But it is very interesting that this 

 Qe cohesion is to be seen in the shepherd's 

 fse, in the wormseed and in the cabbage, as 



II as in the case of the Erucastrum and most 

 the other observed cases of atavistic bracts. 

 is fact suggests the idea of a common origin 

 ■ these anomalies, and would lead to the 

 pothesis that the original ancestors of the 

 ole family, before losing the bracts, exhibited 

 s peculiar mode of cohesion. 



Bracts and analogous organs afford similar 

 ses of systematic atavism in quite a number 

 other families. Aroids sometimes produce 

 g bracts from various places on their 

 idix, as may be seen in the cultivated 

 jenhouse species, Anthurivm, scherserianum. 



