LI^'^'■EA^' socinxv or loxjjon. 55 



When our theory was proinulgafed it no douhL appeared a hig 

 step to assume that th(3 Angiosi)ernious staineu was derived hy 

 means ot" extreme redaction from a pinnate frond-like microsporo- 

 phyll, such as occurred in BennHtites. JXow we know that in the 

 iieunettitalean line great reduction has ajiparently taken phice 

 in this raemb.-'r. Attention is directed especially to tlie micro- 

 sporophyll of Willlamsomtlla cor>iuta, which may carry only two 

 pairs of synangia (53. p. 119). Using this as a parallelism, it is 

 not an improbable assumption tliat the reduction has been carried 

 a stage or two further in the Angiospermous line, resulting in 

 the stereotyped stamen with its pair of bilocular synangia. 



In our paper attention was drawn to the importance of the fact 

 that in the Magnoliacea^ the connective is prolonged beyond the 

 anther as a sterile ti]) (5. p. 48). This vestige, as we believed it 

 to be, suggests comparison with the sterile pointed extremity^ of 

 the Beniiettitean microsporophyll. At any rate this protrusion 

 of the connective beyond the anther may point to the fact that in 

 the Angiospermous line of descent the anther had not originally 

 a!i apical position, but that in the course of evolution it has been 

 left so through the sterile terminal portion of the stamen 

 aborting. 



The additional knowledge acquired in recent years anent the 

 Bennettitalean fructifications has not enlightened us in the least 

 as to the true nature of the female part of the cone. Tliis 

 structure is built up essentially the same in all forms. The 

 simplest view to take is to regard the interserainal scale and seed- 

 pedicel as liomologous, and this is .the one tiie writer at present is 

 inclined to favour (45, & 53. p. 139). It is a " far cry " from the 

 seed-bearing frond of the Pceridosperm to the seed- pedicel of 

 Bennettltcs ; but keeping in mind the great reduction which has 

 apparently taken place in the corresponding male frond, it is not 

 im[)ossible that this pedicel may represent the female frond 

 reduced to its lowest term, viz., to a single stalked ovule. 



Gnetahs. One of the merits of the strobilus theory is apparent 

 in the fact that it finds provisionally a resting-place for this 

 puzzling group. We regarded them as a much niodiiied remnant 

 of an assemblage of plants which left the main stem before this 

 had reached the Angiospermous level. A number of pai)ers 

 dealing with the Grnetales have appeared since, notably from the- 

 pen of the late Prof. Pearson. After a careful perusal of these 

 1 fail to find any new facts which render our standpoint untenable, 

 liecent work tends rather to accentuate iheir relationshij) to the 

 IJennettitales, and to bring them lu^arer to this fossil group than 

 we were inclined to do (c/. 11, 43, 52). 



The male Hovver — morphologically hcrmajihroilite — of IVehvit- 

 scJiia seriously interferes with the endeavour to derive the 

 Gnetales directly from the Conifers. Thompson (55) tries this 

 once t.iore, but limls this flower inconveniently in the way. 

 Likewise attempts to connect the (Inetales witli the Amentales 

 through riroal structure or with the Piperales through sujjposod 



