1894.] MICROSCOPICAL JOURNAL. 333 



by the lower plants in their sexual methods. This shows 

 that the process is practically constant. The embryo is 

 perhaps the most variable of the internal parts of the 

 seed because it comes in contact with the environment 

 after germination, and on the other hand the seed coats 

 are the most constant. Characters drawn from the in- 

 ternal anatomy of the seed should then be those most 

 suitable for the separation of the larger groups of the 

 Phanerogams. 



For the purpose of classification, the first groups into 

 which plants are divided must be general, comprising a 

 large number of forms. The distinguishing characters 

 must, therefore, be common to a large number of species. 

 According to our modern ideas of evolution, these must 

 have existed in the progenator of the group, and have 

 remained nearly unchanged in all of its descendents up 

 to the present time. In species and genera, less general 

 groups, the characters may be drawn from parts less 

 constant. The value of a character then depends on its 

 susceptibility to variation. 



This idea of the value of the internal features of the 

 reproductive organs in the characterization of larger 

 groups is not entirely new. It has already been acknowl- 

 edged by most animal embryologists, and in botany the 

 two larger groups of the Phanerogams are separated by 

 features of the embryo. 



From this standpoint the present discussion of the 

 fruit of the RanunculacesB was prepared. 



The fruits of the Ranunculacese are mostly either de- 

 hiscent several seeded pods or achenes. The achene 

 seems to be the more recent production since the line of 

 dehiscence is still plainly visible between the placental 

 strands of most species, and then too the peculiar struct- 

 ure of the achene with suspended ovules in the genus 

 Anemone lead to this conclusion. Indehiscent fruits are 

 an advantage to the plant only when one seeded, since ii 



