384 BOTANICAL GAZETTE [NOVEMBER 
account of the various characters upon which the systematists 
separate its species. Indeed it appears to be too hard for the 
systematist himself. Scarcely any two systematists agree as to 
which are and which are not true species. This is especially well 
shown by the large number of synonyms which one meets in any 
systematic work upon the genus. 
However, I have found that sterility of pollen is a common 
characteristic throughout the entire genus. But few of the species 
indicate a perfect pollen condition and some are almost entirely 
sterile. In other words, the genus Rubus indicates to me a con- 
dition which appears to prevail in many angiosperms. JEFFREY (21) 
has already pointed out how prevalent sterile pollen is throughout 
the whole family of the Onagraceae, and other groups of plants 
are now being worked upon in this laboratory which seem to show 
a similar condition. 
In closing, it seems to me that many of the species of the 
angiosperms are in a very different condition from that of others 
of the same group and from practically all those of the gymno- 
sperms. They appear often to be natural hybrids which are formed, 
as BRAINERD has pointed out, as a result of a cross between dis- 
tinctly separate species. Such crosses do not result in the segre- 
gation which Mendelian crosses usually exhibit, but rather in the 
blending of the characters of the parents. CoRRENS (31) in his 
observations has noted such a situation in regard to certain plants. 
Such offspring may breed true to their respective characters and 
thus perpetuate a distinct form which, from a systematic point of 
view, is entirely of specific rank, but which should be treated by 
the plant breeder in an entirely different manner. Such forms 
usually have a large percentage of sterile pollen. Thus Rubus 
hybrids formed by distinct species are very sterile, in fact almost 
entirely so, while crosses of varieties which are closely related, 
and which hence are usually more compatible, may give almost no 
indication of their hybrid origin. 
Thus it appears to me that the genus Rubus is one of those 
angiospermic genera in which the term “species” must be used in a 
different sense from that applied to many other members of the 
angiosperms, and practically to all of the gymnosperms. It is a 
