EMBRYOLOGY IN RELATION TO TAXONOMY 



367 



under the Euphorbiaceae. Pax and Hoffman (1931) are in general 

 agreement with Baillon's views but consider that the best course is 

 to assign it to a separate family Callitrichaceae, placed close to the 

 Euphorbiaceae. 



The work of J0rgensen (1923, 1925) has, however, revealed a 

 combination of embryological characters in Callitriche which make 



E B C D 



Fig. 201. Some stages in development of embryo sac and endosperm of Callitriche. 

 A, l.s. young ovule, showing reduced nucellus, single integument, and row of three 

 cells comprising two megaspores and binucleate dyad cell. B, twelve-celled stage 

 of endosperm; only six cells seen in section. C, later stage, showing laying down of 

 micropylar and chalazal haustoria. D,E, portions of embryo sacs, showing rnicro- 

 pylar and chalazal haustoria. (After Jtfrgensen, 1923.) 



all of these assignments seem very unlikely. The ovule is tenui- 

 nucellate and has a single massive integument (Fig. 201 A); the 

 endosperm is cellular (Fig. 201B) and forms well-developed haus- 

 toria (Fig. 201C-E). These features are so characteristic of the 

 Tubiflorae that they suggest a closer relationship of the Callitrich- 

 aceae with the Labiatae or Verbenaceae than with any of the families 

 named in the preceding paragraph. 6 In the absence of adequate 



• This view also receives support from the structure of the fruit and the gyno- 

 basic style of CaUiiriche. 



