4.'J4> THIRD REPORT — 1833. 



longs another part, which is a transformed bud. The stamens 

 are now, according to Professor Agardh, the buds of the leaves 

 of the flower, or of the petals and sepals. They are therefore 

 situated in their axillas, and each stamen belongs to a certain 

 petal or sepal, and both organs together form a little flosculus 

 as part of the whole flower, in the same way as the carpella are 

 parts of the fruit. Thus a Decandrous plant, for example a Ce- 

 rastium, consists of 10 floscules, each consisting of a leaflet and 

 of a stamen. A Pentandrous plant, for example a Borago, 

 consists also of 10 floscules ; but 5 of them, those of the interior 

 verticillus, are incomplete, bearing no stamens in their axilla?. 



The observations made on the situation of the parts of the 

 flower wei-e collected by Professor Agardh into the following- 

 general laws or views. 



1. The number of stamens, in all cases where this num- 

 ber is determinate, depends upon the number of sepals and 

 petals ; and when there seems to be a different normal num- 

 ber for the leaves of the flower and for the stamens, it is an 

 aberration arising from abortion. 



2. The difference between the flowers which have the same 

 number of stamens as of leaflets, and those where their num- 

 ber is only the half of the leaflets, is caused by the abor- 

 tion of the stamens of a whole verticillus of floscules ; and ge- 

 nerally of the corolline floscules. The reason is, that the corol- 

 line verticillus is constituted of later parts, which do not arrive 

 at complete development. 



3. The same reason is to be assigned to the general fact that 

 the petaline stamens are generally the weaker, smaller, and 

 later. 



4. The determinate stamens are either 1, 3, or 5 in num- 

 ber, belonging to each leaflet, because the buds, according to 

 Professor Agardh's theory laid down in his above-mentioned 

 work, originate properly in the axillae of two deviating fasciculi 

 of spiral vessels, which fasciculi in the leaf (being no other 

 than the nerve) are always 1, 3, or 5, &c., and the bvids must 

 therefore have the same symmetry and number. This is the 

 reason of the determinate number of stamens in several Polyan- 

 drous families, as, for example, in Rosacece, in which the sepals 

 have each 3 stamens, and the petals each 1 ; and in Phila- 

 delphice, in which the sepals have 5 stamens, and the petals 

 1 : whence the former family must have 20 stamens, and the 

 latter 24. 



5. Some Polyandrous plants have not a determined number 

 of stamens. In these Professor Agardh distinguished two cases. 



