of Variation in the Corolla of Veronica Buxbaumii. 81 



an oblique instead of an upright cross. The morphology of this 

 flower is not quite clear, but there seems little doubt that it 

 should be regarded as a corolla in which the posterior petal has 

 been divided as in Fig. 2, while the anterior and lateral elements 

 have developed in the same way as those shewn in Fig. 4. 



Occasional forms. Besides the forms enumerated above, of 

 which all are found in some quantity, there are various other 

 forms of only occasional occurrence. Of these some are sym- 

 metrical, e.g. corollas having three posterior petals making six in 

 all (Fig. 11); corollas having 7 or 8 petals, viz. 3 or 4 anterior, 

 3 or 4 posterior (Fig. 10). There are besides these asymmetrical 

 forms of variation, in which for example one of the lateral petals 

 is smaller than the other, and so forth, but none of these are found 

 in any quantity. 



Small flowers. In the previous paper mention was made of 

 a peculiar form of flower having its parts, especially the corolla, 

 of very small size (Fig. 18). These flowers were of occasional 

 occurrence only. 



Galycc. The calyx is also liable to numerical variations, but 

 no record of these was kept. There are frequently variations of 

 the calyx in correlation with those of the corolla, but not rarely 

 the one whorl may be abnormal while the other is normal. 



Stamens and Pistil. Variations in these parts were extremely 

 rare. 



Method. The plants selected in each year for self-fertilisation 

 were potted and completely isolated, usually by means of a bell- 

 jar over water, or a fine muslin net. Before covering, the flowers 

 then open were all cut off, and the artificial self-fertilisation was 

 not begun until the following day. The self-fertilised flowers 

 were marked with coloured threads and the cover left on the plant 

 until the corollas of these flowers had fallen off. The cover was 

 then removed and the capsules ripened in the open. 



The seeds were sown either in the autumn or in the early 

 spring. 



The seeds from each selected capsule were sown in separate 

 pots, and afterwards pricked out in an open piece of ground in the 

 Botanic Garden. 



The flowers, which open in the morning, were counted during 

 the morning hours on as many days as possible during the flower- 

 ing season. 



Except during the first weeks of 1892 each flower was picked 

 as it was noted, thus all danger of repetition was avoided. 



VOL. X. PT. II. 6 



