LIFE-HISTORY AND PHYSIOLOGY. 



1 1 



a relation to the functions of which they are the channels. 



The Perianth segments vary from half an inch (0.013 Metre) in C. parvifiorus 

 (Plate LXVII) to two and a half inches (0.063 Metre) in length, as in C. speciosus 

 (Plate LX1V); and from an eighth of an inch (0.0032 Metre) to three-quarters of an 

 inch (0.019 Metre) in width. Most commonly obovato-lanceolate in form, but varying 

 much in outline; occasionally emarginate at the summit, but more frequently pro- 

 duced into a point. The form, and also the relative length and width are not 

 constant, and cannot, with -one or two exceptions, be relied on as specific characters; 

 and an emarginate and finely pointed segment may occur in the same flower. In 

 C. vallicola, (Plate II, figs. 1 and 13) however, the curious acuminate segment is a 

 well marked and constant feature. The three inner segments are always somewhat 

 shorter than those of the outer whorl: in C. iridiflorm (Plate I,) the difference is so 

 much more striking than in any other species that it suggested to Schur its generic 

 separation as Crociris. 



There appear to be two distinct sets of colour cells in the segments, the 

 disposition of colour on the inner surface never exactly corresponding with that of 

 the outer. By a little dextrous manipulation the cells of the segments can be 

 peeled off as three distinct layers, the inner or middle layer being almost colourless. 

 Excepting in the self-coloured species, the colouring of the outer surface of the 

 outer segments is notably different to that of the inner surface, as well as to both 

 surfaces of the inner segments. In the species and varieties with distinctly striped 

 and feathered markings, the feathering is confined to the outer surface of the outer 

 segments, excepting in C. versicolor (Plate XVI), in which the feathered markings 

 are nearly uniform on all six segments. In all other species the inner whorl is 

 either differently feathered, or unfeathered with just an indication at the base of a 

 few lines of incipient feathering. 



Although the feathered markings are usually confined to the outer surface 

 of the outer segments, this is not invariable, and in three or four species, namely, 

 C. zonatus (Plate IV), C. vallicola (Plate II), and C. pulchellus (Plate LXV), the 

 purple markings are on the inner surface. C. pulchellus also has separate markings 

 externally. 



In the feathered blue and white species the featherings are purple; in the 

 orange species, of a rich bronze-colour. But in each case the colour-cells of the 

 markings are purple; the bronze colour of the feathering of the orange species 

 being due to the overlay of purple on the orange ground; and in the pale forms 

 and albinos of the orange species, the markings are developed as purple instead 

 of bronze. 



Feathered markings are not constant as a specific character, as the majority 

 of species vary with self-coloured and feathered flowers. Indeed, colour itself is of 



