docs not (lifter Ironi .\.\ in llio kind of clironiosouK'S, l)ul ()nl\' in 

 having two sets. They are thus hke Oenothera (jigos, and are 

 found to exhibit a marked increase in size of all their parts. In 

 certain cases triploids, as AAA, have been found among garden 

 varieties. Hurst remarks that they may have arisen "from a 

 duplicate gamete AA arising in a diploid AA, or from a cross be- 

 tween diploid AA and tetraploid AAA A, or from a bud six)rt 

 derived from a somatic cell of a tetraploid AAAA that had lost a 

 set of chromosomes." 



The dilTerential jxjlyploids are those containing more than one 

 kind of septet of chromosomes; these are divided into regular and 

 irregular. Thus Rosa Jiuntii Hurst (a new species from China) 

 is AABB. Rosa ceiitifolia L. is AACC. Rosa palustris Marsh 

 is AADD. Rosa davidi Crep. from China is AAEE. Rosa spino- 

 sissi))ia L. is BBCC. Rosa pirn pinell. folia L. is BBDD. Rosa 

 multibracteafa Hemsl. & Wils. from China is BBEE. Rosaznrgin- 

 iana Mill, is CCDD, as also are R. suffidta Greene, R. lunelli 

 Greene and others. No CCEE species is yet known. R. pendtdina L. 

 is DDEE. In addition to these tetraploids, there are hexaploids, as 

 the beautiful R. moyesii Hemsl. & Wils. (AABBEE), which I 

 saw growing at Cambridge; R. zmlsoni Borr. (AABBCC) of 

 Wales, Ireland and Scotland ; the North American R. nutkana 

 Presl. (AADDEE) andR. eng el manni Wats. (BBDDEE). There 

 are even a few octoploid species, namely R. tackhohnii Hurst n. sp. 

 (AABBCCDD) from Iceland, and R. acicidaris Lindl. 

 (BBCCDDEE), circumpolar in subarctic regions. 



The remaining group contains the irregular septet species, in 

 which only part of the groups have synaptic mates. Such species 

 are confined in nature to temperate Europe and Western Asia in 

 an area approximately equivalent to that covered by the Pleisto- 

 cene glaciation. Apparently they do not occur as far east as Lake 

 Baikal ; at least I believe none of the material I obtained there was 

 so referable, though the results of Hurst's examination of it have 

 not yet been reported. "All these species present the phenomenon 

 at present unique in plants and animals, of a regular but unequal 

 reduction division in female gametogenesis, which causes them 

 to produce female gametes carrying from two to five times as many 

 septets of chromosomes as their male gametes, so that their recipro- 

 cal hybrids are entirely different in their chromosome content and 



