370 THE HIGHER DICOTYLEDONS. 



the length of the head-stalk, (2) the shape of the expanded 

 top of the stalk (which is at first flat], (3) the size of the 

 ovary, (4) the length of the slender stalk (at first very short) 

 just above the ovary, (5) the ring of hairs at the base of the 

 (now withered) corolla-tube? What becomes of the upper 

 parts of the flower (corolla, stamens, style) ? When the 

 fruits are ripe, the stalk becomes erect again. Does the head 

 still open on fine days, or does it remain closed and for how 

 many days ? How is the change in form of the stalk -top 

 connected with the change in form of the whole head ? The 

 " disc " is now becoming rounded (convex) instead of flat, 

 while the fruiting head becomes a globular mass about two 

 inches in diameter ; the flat fruits require, therefore, to be 

 spaced out on the " disc." 



Do all the flowers in a head grow into fruits? Do you 

 notice any difference in this respect between plants growing 

 in sunny places and plants growing in shaded places ? Why 

 do the latter usually produce a smaller proportion of fruits 

 than the former ? Why are most of the unfertilised flowers 

 usually found at the outside of the head ? Probably the last 

 three questions can be answered after considering the facts 

 that efficient insect- visits are far more frequent (1) in sunny 

 places and (2) in the more fully expanded and therefore 

 more conspicuous heads, i.e. after the first day on which the 

 heads are open. 



These observations raise the question, is self-pollination an 

 effective substitute for cross -pollination ? I have tried ex- 

 periments, designed to prevent insect- visits, with rather 

 variable results, but on the whole the output of normal fruits 

 has been much smaller than in neighbouring heads left un- 

 touched. Tie small bags of muslin or, better, paper over young 

 unopened heads, and see how many fully developed fruits 

 are produced. In watching Dandelion-heads and noting all 

 that happened day by day I have found that the average 

 time from the first opening of the flowering head to the final 

 opening of the fruiting head (with fruits ready for dispersal) 

 is about eighteen days ; it is lengthened a little by rainy 

 days, shortened when the weather is fine. However, in this, 

 as in all other readily verified matters, make your own 

 observations instead of merely accepting those of other 

 people. 



