30 REVIEW OF THE GENUS NAECTSSUS. 



Corona. — The crown in the centre of the flower varies very much, 

 and furnishes some of the best marks which we can get for character- 

 izing the species. Beginning with N. Broussonetii, it shows itself as 

 a hiyer ahnost entirely confluent with the upper part of the tube. We 

 get it in the next stage of development in the autumn-flowering spe- 

 cies, in which it is an erect rim, scarcely over half a line high. In N. 

 posticus and N. bifloriis it is a similar but rather deeper rim, with a 

 very much crisped scariose edge. In the N. Tazetta group it is a cup, 

 from a third to a quarter as long as the divisions of the perianth. In 

 N. incomparabilis and N. odoriis it is an erect cylinder, half an inch 

 deep, and quite as broad as deep, whilst in N. Pseudo-Narcissus and 

 N. Bnlbocodium we have it either as a reversed cone or a cylinder with 

 a spreading mouth, which is quite as long or rather longer than the 

 divisions of the perianth. 



Stamens. — Usually the length of the filaments is in proportion to 

 that of the corona. In N. Tazetta, N. poeticus, and most of the 

 others with a short crown, the stamens are nearly or quite sessile in 

 two contiguous rows near the throat of the tube ; but we have a 

 marked exception to this in JV. Broussonetii, which has filaments one 

 and a half to two lines long. In N. odorus, N. incomparabilis, N. 

 montanus, and N. Macleaii, the filaments about equal the anthers. In 

 N. calathinus and N. triandrus they are two or three times as long as 

 the anthers, the lower three issuing from near the base, and the upper 

 three from near the top of the tube. In N. Pseudo-Narcissus they all 

 spring from one level low down in the tube, and are quite straight and 

 more than half an inch long. In N. Bnlbocodium they spring from 

 near the base of the tube, and reach nearly or quite to the mouth of 

 the crown, and are all distinctly upcurved toward the point. 



Style. — Herbert attaches great importance to the length of the style 

 in characterizing species and varieties, but it certainly is not at all safe 

 to do so. In general terms I may say that the stigma is on a level 

 with the upper row of stamens. 



Colour of the Flotoer. — In nearly all the species there is consider- 

 able variation here. Tlie corona is more constant in colour than the 

 rest of the flower, and generally deeper in hue. From snow-white 

 and milk-white, we pass gradually to a bright yellow, getting no other 

 bright shade of colour except green in the limb of one rare, little- 

 known species, and bright red in the crown of N. poeticus. 



