164 Scientific Proceedings, Royal Dublin Society. 



Discussion of Results of Table III. 



Each of tlie above five varieties of /. Xiphium showed a sliarply defined 

 inhibition-area, which remained colourless when treated with tlie usual 

 reagents. This area, on the haft of the falls, coincided with the distribution 

 of the plastid yellow. Prolonged treatment with the reagents in part 

 obliterated this white area, apparently because the inhibitor slowly diffused 

 away. 



As before, restoration of colour took place in the deeply coloured forms. 

 This group is apparently characterized by theabsenceofiniiibitor, except over 

 tlie sharply defined area referred to above. A similar inhibition has been 

 noticed by Keeble and Armstrong (4) over the yellow eye of Primula 

 sinensis. The strong reaction witli benzidine of the "Thunderbolt" and 

 deep-blue (No. 14) forms is noteworthy, as it brings out the relation between 

 oxydase action and intensity of natural pigmentation wlien undisturbed by 

 inhibitors. 



The absence of a-naphthol reaction in " Tiiunderbolt " epidermis agrees 

 with what was previously noted. With the blue epidermis of No. 14 the 

 reagent reacts strongly ; here it seems as if a blue chromogen does not 

 inhibit the reaction, while a purple does. 



The yellow in " Thunderbolt," producing the bronzed shade, is probably 

 a sap soluble yellow; for, on treatment witli strong spirit for some hours, 

 the falls were almost indistinguishable from those of tlie deep-blue No. 14, 

 only the yellow on the haft remaining, as it had not all been dissolved out of 

 the plastids ; furthermore, the inhibition area coincided with the plastid 

 distribution in this group. Direct observation must settle this point. 



" Blanche superbe," having neither a sap-pigment nor inhibitor, would 

 probably behave as a Mendelian recessive if crossed with an anthocyan- 

 containing flower. The same applies to " chrysolora." In No. 13 the absence 

 of chromogen is most likely the deciding factor, for there is a pronounced 

 oxydase reaction with both reagents. 



Table IV contains the results obtained with the Pogoniris group. 



