ward typical synapsis stage in which the chromatin threads are 

 often seen to be arranged parallel to each other and later fuse 

 and become one or several thick threads. The heterotypic spindle 

 figure deviates somewhat from its correspondent in H. auricula, 

 as the chromosomes are significantly rich in numbers. They are 

 different in shape and form, but could in any case be classified in 

 bivalent and univalent chromosomes, very much reminding one 

 of the Dr oser a-hyhrid, formerly described by me (Rosenberg 22). 

 Figs. 15 and IV A will show more clearly the point in question. On 

 account of the great number of chromosomes it is very difficult to 

 determine their number with absolute accuracy. I have often 

 found rather regularly 14 or 15 bivalent and 6 or 7 univalent 

 chromosomes, whilst in other cases about 17 entirely bivalent 

 chromosomes could be seen, 

 can correctly value these very 

 various spindle figures. In 

 the beginning I found reasons 

 which caused me to suppose 

 that a hybrid from, two pa- 

 rents with a different number 

 of chromosomes, had to be 

 reckoned with, but later I 

 observed that my attention 

 had to bê given to a sort 

 of division which could be 

 regarded as an intermediate stage between reduction division and 

 vegetative division. In the aforementioned Drosera-hybvlå always 

 20 bivalent and 10 univalent chromosomes could be recognized, 

 whilst, in this case a very great variation in the proportion of 

 bivalent and univalent chromosomes could be found. It therefore 

 seems to me very probable that in H. excellens there is an incom- 

 plete reduction process. I should like to express it in the following 

 manner: the great number of chromosomes in each gonomer still 

 keeps its „ affinity" and therefore can be united with its corre- 

 spondent in the other, whilst a part, different in different nuclei, 

 have lost that characteristic and were therefore univalent. 



It seems to me that already in the synaptic and postsynaptic 

 stages such incomplete reduction processes are to be observed. 

 The fig. 14 represents for instance a stage soon after synapsis 

 where among the double chromatin threads also single ones can 



I have^ long doubted as to how we 



A B 



Fig. IV. H. excellens, A, heterotypic spindle ; 

 B, anaphasis in the second division, with free 

 chromosomes between the daughter nuclei. 



