356 Mr riiilip Sewell's Observations upon the 



Wc shall later refer in more detail to the seeming relation 

 between general habit, together with the nature of the young 

 leaves, to the percentages of successes and failures to 

 produce healthy plants ; at present we may point out some 

 more evident facts regarding distril^ution and germination, 

 which are not, however, of much value apparently. 



Thus, the first four species to germinate are natives of the 

 South of Europe, Asia Minor, or the Canaries, where, doubt- 

 less, the conditions of climate in the open air are very 

 similar to those at La Mortola, Again, the species which 

 have not germinated at all are from Brazil, New Granada, 

 and (actually !) Italy. The fact that >S^. Dominica (grown 

 under that name, and in another pot under the synonym 

 S. gravcolcns) should not have germinated, although the 

 experiment has been made in the country where the plant is 

 indigenous, is just one of those facts which shew us of how 

 little value it may be to judge as to a plant's requirements, 

 without reference to exact habitat or to the nature of the plant 

 itself. It is, of course, possible that the seed of this species 

 has been faulty, but more likely the conditions under which 

 it was attempted to germinate it have been different from 

 those of its habitat ; perhaps an overplus of watering may 

 afford an explanation. We imagine also that we have 

 detected a connection between the nature of the coat of the 

 nutlet and the amount of success in germination. It appears 

 as if those of smaller size, with shining or smoother coats, 

 had germinated less readily ; certainly less successfully, as 

 measured by the total percentage out of the fifty or one 

 hundred seeds sown in each case. 



This, which we put forward as little else than a specula- 

 tion at present, does not seem so unnatural to us when we 

 think of the vastly different habitats — the margin of a lake 

 or the slope of a hill-side, where the species naturally scatter 

 their seeds. 



Yet we may point out that, although in the same species 

 when germinated under the different synonyms of S. 

 Tenorii and >S'. variegata, there was close agreement in 

 response to the environment in which they were placed for 

 germination, yet in the two forms of S. interrupta there 

 was a difference of forty days in the time of germination. 



Similarly, confining our attention to the percentages which 



