81 Chronicles of Science. [Jan., 



character in distinguishing the species. He thinks that the pappus 

 may thus furnish distinctive marks in the case of Sonchus arvensis, 

 arboreus, asper, oleraceus, palustris, and tenerrimus. All these species, 

 and especially the last two, show best this system of recurved teeth, 

 whilst the pappus of Sonchus divaricatus is markedly different, and 

 approaches near to the species S. Eabdotheca. 



A recent monstrosity observed in Anagallis arvensis seems to show 

 distinctly a placenta formed from the axis, and the outer integument 

 of the ovule formed from a leaf. The axile bud was formed from the 

 sepals and petals, as well as from the leaflets forming the ovary. 

 The specimens showed very completely the foliaceous nature of the 

 organs. 



Professor Cramer, from numerous researches into monstrous ovules 

 and into the development of normal ovules, maintains that those of 

 JPrimulacece, Composite, &c, are metamorphosed leaves, and those of 

 Leguminosce, Umbelliferai, Banunculacece, &c, are metamorphosed parts 

 of leaves. 



The normal tetracarpellary condition of the siliqua in Cruciferce 

 has been recently advocated. Four carpels are actually found in 

 Tetrapoma pyriforme. The normal number of stamens has been con- 

 sidered eight or sixteen — the latter number occurring in Megacarpcea 

 polyandra. 



Dr. Alexander Dickson has given the results of his investigations on 

 Malope and Kitaibelia. He finds that the development of the androe- 

 cium, as regards the evolution of the staminal lobes, is essentially the 

 same in both genera. The androecium at first appears in both as an 

 even rim-like cushion, pentagonal externally, surrounding the flattened 

 termination of the floral axis, which appears as a pentagonal depression. 

 The external angles of the staminal cushion in both genera alternate 

 with the sepals. The essential difference between the arrangements in 

 the two genera consists in this — that in Malope, the five angles of the 

 central depression are superposed to the external angles of the cushion ; 

 that is to say, the sides of the depression are parallel to the outer sides 

 of the staminal cushion ; while in Kitaibelia, on the other hand, the 

 angles of the central depression alternate with the external angles of 

 the cushion, the sides of the depression thus lying crosswise to the 

 outer sides of the cushion. This is seen long before there is any 

 appearance of the carpels. In both genera five pairs of lobes alternate 

 with the sepals, and extending longitudinally in a radial direction are de- 

 veloped on the surface of the cushion, which, as development advances, 

 grows up as the staminal tube. In the subsequent stages, the author 

 fully confirmed the statements of Payer as to the development of the 

 stamens. On each of the lobes a longitudinal series of mammilla} appears 

 in a centrifugal succession, and from above downwards. Later, each 

 of these mammillae branches into two lobes, each of which is developed 

 into a stamen, with an ultimately one-celled anther ; so that at last 

 there are twenty rows of stamens, or, perhaps, more correctly, two rows 

 of bifid ones. The carpellary groups form the sides of a pentagon, 

 which in each genus corresponds to the sides of the original central 



