SOMATIC CHROMOSOMES IN VICIA 3l5 



ting are ail going on simultaneously ; only in a given small portion of a 

 thread do they follow in definite séquence. Furthermore, as soon as the 

 threads become equalized they at once begin to thicken, so that when the 

 vacuolization and splitting are a little delayed they occur in threads of 

 greater thickness than those of fig. lO (compare fig. 11-16). In fig. 12 

 the split is complète while the threads are yet very thin, while in the nu- 

 cleus of FIG. 13 sevcral much thicker threads, as at a, are still in the vacu- 

 olate stage. The variation in thickness shown by différent threads of the 

 same nucleus is also noticeable, fig. 15, 16. 



Ail steps intermediate between vacuolization and a complète split are 

 easily made ont in this séries ot figures. In fig. î4 are given several in- 

 structive détails from différent nuclei. In the left hand portion of the chro- 

 mosome e the split is still in the stage of vacuoles and open spaces, the 

 cross pièces (originally the parts between the vacuoles) being heavy and 

 numerous. In the right hand portion of the chromos'jme most of the cross 

 pièces hâve disappeared, leaving the two halves almost entirely free from 

 each other. Fig. 14 a, /', d show further examples of the cross pièces still 

 uniting the halves of the chromosome. Fig. 14c represents a very excep- 

 tional condition, the split being in the vacuolate stage in a very much 

 thickened thread. Other chromosomes in the same nucleus show a com- 

 plète split. Since the phenomenon was observed in this one case only it is 

 to be regarded as a variation without spécial meaning. It only serves to 

 emphasize the prophasic nature of the longitudinal split. 



In the présent investigation of Vicia there is found no trustworthy 

 évidence in support of the view of Strasburger (07, 11) that the splitting 

 is initiated by the division of chromatic units or ^ chromomeres ^ arranged 

 in a séries along the thread at prophase. The modified opinion of Mûller 

 (12), who holds that the portions of the thread between the chromomeres 

 split first, is in better accord with the aspect of the threads as they begin 

 to split, but that the splitting is bringing about the division of definite 

 morphological units, as both authors believe, we are not able to conclude 

 from the évidence at hand. This subject will be dealt with more fully in 

 the gênerai discussion. 



x^s the double threads thicken they also become more uniform in size 

 throughout their length, fig. 17. In this nucleus the}^ hâve a gênerai lon- 

 gitudinal orientation which makcs them especially favorable for study. The 

 halves lie very close together at this time, but the split is easily seen in ail 



