194 THE SCOTTISH BOTANICAL REVIEW 



flowers no nectar is formed, there is no "concrescence" (in 

 reality it is only agglutination or sticking together) of the 

 anthers, and the pollen is formed in great quantity, as is the 

 case with other anemophilous flowers. The condition of the 

 exine indicates, however, that originally pollination took 

 place by insects as in other Compositas. 



It is well known that the pollen in entomophilous plants is 

 provided with a spiny or sticky exine, whereas in anemo- 

 philous plants the exine of the pollen-grains is smooth and 

 not sticky. The Ambrosiaceae, so far as I have been able to 

 examine them, still show in the exine the remains of spinose 

 thickenings such as are found in numerous other Compositae, 

 but these are so reduced that the pollen is no longer adapted 

 for adhering. 



The male flowers still show on the rudiment of the stigma 

 a hairiness which recalls the " stigmatic brush" of other 

 Compositae. In the female flowers, in which the stamens 

 have dwindled without leaving a trace, there is no longer 

 any indication of this stigmatic brush. Thus the Ambrosiaceae, 

 in contrast to other Compositae, are retrogressive in structure. 

 On the other hand,' it is evident that they have developed new 

 characters not attained by other Compositae. This is seen 

 more especially in the character of the noteworthy envelopes 

 with which the fruit is enclosed. The fruiting heads of 

 XanthitiiH are well enough known to the wool-grower as the 

 detested "burrs" (Kletten), which by means of their bent 

 hooks attach themselves to the woolly coats of animals, and 

 are so dispersed. We shall see that as regards these fruiting 

 heads the other Ambrosiaceae have also undergone retrogres- 

 sion in contrast with Xanthium. 



It is not surprising that such noteworthy plants have 

 repeatedly formed subjects for investigation. So far as the 

 developmental history is concerned, and apart from Payer's^ 

 contributions, which in the present instance need only be 

 indirectly considered, there are available only the investiga- 

 tions of Rostowzew,^ and these I propose in the following to 

 supplement on some points. 



^ Payer, " Traite d'Organogenie comparee de la fieur " (Paris, 1857, p. 638). 

 ^ T. Rostowzew, "Die Entwicklung der Bliiten und des Blutenstandes bei 

 einigen Arten der Gruppe Ambrosiese" (" Bibliotheca botanica," 20). 



