132 BOTANICAL GAZETTE [FEBRUARY 
’ the micropylar haustorium arises from a single cell and contains 
four nuclei. 
ScHMip has made it clear that the different species of a genus 
closely resemble one another in the character of the endosperm and 
haustoria, and he believes that these characters should receive con- 
sideration in drawing up a natural classification of the family. 
Judging things from this point of view, the Rhinantheae appear to 
be a natural group, since all the members studied show an extraor- 
dinary similarity of development. It certainly is striking that 
Linaria and Striga, which are widely separated in the present system 
of classification, should show such close agreement in their develop- 
ment of endosperm and haustoria. Though ScHMipD has not paid 
much attention to the proembryo, it seems that here also is a strong 
resemblance. Surely there must be a close relationship between 
genera which show agreement in minute details of the development 
of organs which do not seem to be influenced by environment to 
nearly the same extent as those organs which are most used in 
drawing up a scheme of classification. 
Dop (7) finds in Buddleia a remarkably close resemblance to 
Verbascum, Scrophularia, and Digitalis.as regards the mode of 
formation of the endosperm and haustoria. He concludes that the 
evidence obtained emphasizes the relationship between the Scroph- 
- ulariaceae and Buddleiaceae (if the latter be regarded as @ 
family distinct from the Loganiaceae as WeETTSTEIN [16] thinks it 
ought to be), and that it tends to separate the Buddleiaceae more 
widely from the Solanaceae. 
The question whether parasitism of the plant as a whole affects 
the embryo sac and embryo has often been raised. The evidence 
afforded by the Scrophulariaceae all tends to show that these 
structures are not influenced by the habit of the plant. Extensive 
development of the haustoria is certainly characteristic of 
the Rhinantheae, but Striga, which is also a semi-parasite, 
is not so characterized, while Veronica, a non-parasite, has 
practically the same degree of haustorial development as the 
Rhinantheae. It seems, therefore, that intensity of haustorial 
development is not to be correlated with the parasitic habit of 
the plant. 
