OP THE STAMENS, ETC. 29 



normal circumstances, the petals beconae fused together 

 by their edges along their whole extent, at the base 

 only, at the apex only, as in the Vine, or at the base 

 and apex, leaving the central portions detached. In- 

 dications of the junction of the petals may generally 

 be traced by the arrangement of the veins, or by the 

 notches or lobes left by imperfect union. In Crocuses 

 I have frequently met with cohesion of the segments 

 of the perianth, by means of their surfaces, but the 

 union was confined to the centre of the segment, 

 leaving the rest of the surfaces free. 



Cohesion of the stamens. Under natural circumstances, 

 cohesion of the stamens is said to take place either by 

 the union of their filaments, so as to form one, two, 

 or more parcels (Monadelphia, Diadelphia, Poly- 

 adelphia) ; at other times, by the cohesion of the 

 anthers (Syngenesia), in which latter case the union is 

 generally very slight. It must be remembered, how- 

 ever, that the so-called cohesion of the filaments is in 

 many cases due rather to the formation of compound 

 stamens, i. e. to the formation from one original staminal 

 tubercle of numerous secondary ones, so that the 

 process is rather one of over development than of 

 fusion or of disjunction. These conditions may be met 

 with as accidental occurrences in plants or in flowers, 

 not usually showing this arrangement. Thus, for 

 instance. Professor Andersson, of Stockholm, describes 

 a monstrosity of Salix cahjculatay in which the stamens 

 were so united together as to form a tube open at the 

 top like a follicle.^ This is an exaggerated degree of 

 that fusion which exists normally in Salix inonandra, 

 in Cucurbits and other plants. 



Cohesion of the pistils is also of very frequent oc- 

 currence in plants, under ordinary circumstances, but 

 is less commonly met with than might have been ex- 

 pected as a t^ratological phenomenon. 



' Joamal of the Linn. Soc. Bot.,' vol. iv, p. 55. 



