Botany. 81 



wall of the ovary, through a distinct digestive process, and thus 

 the ovule possesses only a transitory existence as such. 



The vast majority of Phanerogams are, of course, Seminees. 

 Under the head of Bitegminate in this class come all the Gymno- 

 sperms and most of the Gamopetalous Dicotyledons. The 

 Bitegminate comprise all the Monocotyledons, with the excep- 

 tion of Graminew, and most Dialypetalous and Apetalous Dico- 

 tyledons. 



The primary group of Inseminees is much smaller but much 

 more diversified. The author at this point calls attention to the 

 inexactness of the designation Spermaphytes, and shows that 

 there is no absolute necessity for Phanerogams, as such, to have 

 what we have called seeds. The importance of this group of 

 Inseminees from a systematic point of view has been empha- 

 sized by Professor van Tieghem, in the communications above 

 referred to. 



Of these, the Inovidees, or Loranthaceaz as enlarged, are divided 

 into ten orders, all of which are characterized by the absence of 

 true ovules. It will, perhaps be remembered that his views in 

 regard to certain Loranths, propounded by van Tieghem, in 1869, 

 in his study of Vtscum, and unfavorably criticised at the time, 

 have been confirmed in good degree by Treub and by Johnson. 

 These ten orders are divided into 21 tribes and 141 genera, 

 more than one hundred of which are new to science. 



The Innucellees, or Santalineas (including the Olacales) are, as 

 re-arranged, placed in two alliances, nine families, nine tribes, 

 and fifty genera, of which five are new. 



The Inseminees with a naked nucellus constitute the new sub- 

 division, Anthobolinees^ a single family, with four genera. 



The Unitegminate Inseminees are the Icacinees. The two 

 alliances comprise fifty-two genera, many of which are new. 



The Bitegminate Inseminees have two sections, three alliances, 

 seven families, and more than three hundred genera. In this 

 group the author places the grasses, and notes the wide separa- 

 tion between Graminece and Cyperacece. 



Some idea of the results following the adoption of Professor 

 van Tieghem's system can be gained by a slight comparison of it 

 with two accepted methods of classification. In Bentham and 

 Hooker's Genera Plantarum, the plants which are here called 

 Inseminees are placed in four families, namely loranthaceoe, 

 Santalacew, Olacacew, JSalanophoracew, with about 90 genera. 

 In Engler's treatise, now in course of publication, Myzodendraceai 

 is separated from Santalacew, and Icacinem from Olacacew, mak- 

 ing six families and 120 genera. Van Tieghem makes of these, 

 36 families and 260 genera. Moreover, he states that this differ- 

 entiation cannot stop here : increased knowledge of some of the 

 forms will necessitate still further subdivision. 



In the last paper of the present series, the author considers the 

 Seminees, dividing them first of all into Astigmates and Stig- 

 mates. These correspond very nearly to the Gymnosperms and 

 Angiosperms, or, otherwise, Arehigoniees and Anarchegoniees, or 



Am. Jour. Scl— Fourth Series, Vol. IY, No. 19.— Jul^, 1897. 

 6 



