SOMATIC CHROMOSOMES IN VICIA 321 



tween the chromomeres split first. In igi2 Lundegârdh (12^) denied the 

 existence of chromomeres in Vicia. 



It is clear, then, that we hâve to do hère with the opinion of Mûller, 

 which may in ail probability be regarded as also the last opinion of 

 Strasburger. The theory put forth by Muller has much support in the 

 observed aspects. In Vicia, when the vacuoles along the thread become 

 open spaces, relatively heavy cross pièces are left Connecting the longitu- 

 dinal halves, fig. 15 at a, fig. I4e at left. This stage Mûller shows 

 in his fig. II as a séries of rounded granules joined by two délicate 

 strands. When the cross pièces become thinner or finally break at the 

 middle to complète the split the resuit is the accumulation of material in 

 small lumps opposite each other, fig. 10, 13, i5 and 16 at b, Mùller's 

 fig. 12, which gradually become equalized with the rest of the thread. 



This process we cannot for several reasons interpret as representing a 

 division of morphological units or chromomeres. There is in Vicia no 

 such sharp distinction between rounded granules and Connecting parts as 

 Mûller shows in his fig. 11 and 12c, but the heavier portions pass ofî 

 more gradually into the lighter without any change in the nature of the 

 material. An apparent differentiation of the two is brought about by the 

 more rapid removal of the stain from the délicate parts. The occurrence of 

 such pairs of small swellings is by far too irregular in Vicia to be of impor- 

 tance; we find nothing equal to Mùller's very regular fig. 12a. 



Furthermore, if such granules be ultimate units they are far too few 

 for the présent requirements of mendelian théories, as other writers hâve 

 often objected. If each is held to be an aggregation of invisible primary 

 morphological units, each of which splits and thus brings about the divi- 

 sion of the visible granule, the theory of chromomeres has gone where the 

 microscope cannot at présent follow. We are inclined to remain more 

 closely with the observed facts and regard the split as the division of a 

 homogeneous thread or band, in which there is as yet no trustworthy évi- 

 dence for the existence of chromatic morphological units. At the same time 

 we cannot fail to re'cognize the highly suggestive nature of the arrangement 

 of the chromatin in a thin thread and its accurate séparation into two 

 equal parts. In the absence of direct morphological évidence for the pré- 

 sence of varions » qualities ^ placed in a séries along the thread we may 

 still look hopefuUy toward other sources for the further support which it 

 would seem the theory of Roux must sooner or later hâve. 



42 



