4 ARKIV FÖR BOTANIK. BAND 17. N:0 11. 



six of them, 75 chromosomes could be easily counted; in 

 the sixth only 74 could be found, the missing one being 

 probably hidden by one of the others. Hence we can state 

 with absolute certainty that the somatic chromosome 

 number in the Bafios-race is 7 5. Obviously, this 

 race is highly polyploid, and we may conclude that the 

 bivalents and univalents of the heterotypic division should 



« 





• •*•• ••• 



►%•• 



v^. 



• •• • 



»V •• 



. • • . •• • 



• 1^ •• ••• 



• ••• • •*• • 



• •••• 



Fig. 3. Pollen tetrad of the Bauos-race. 4. Pollen tetrad of the Tenguel- 



race. 5 — 7. Somatic metaphases of the Bauos-race. — Reich, horn. imm. 



Vi2; ok. 4 (Figs, 3-4), Leitz peripl. ok. 20 X (Fig.s. 5—7). 



be found in numbers that represent multiples of a funda- 

 mental chromosome number. From this and from the count- 

 ings, actually made in the pollen mother-cells, only one 

 combination seems conceivable, viz. 30 bivalents and 15 uni- 

 valents. Small irregularities, dependant on varying conjugation, 

 may perhaps occur (e. g. 29 biv. and 17 univ. etc.), but still 

 we are certainly justified in concluding that this Ananas- 

 race must be a hybrid between two other races 

 with 30 and 45 as haploid chromosome numbers. 

 The fundamental number of the genus is probably 15 or 5. 



