iv PREFACE. 



Eriocaulon, in which the species are separated by- 

 characters which can be determined only with the help 

 of a good lens, some of the flowers being under a twelfth 

 of an inch. I doubt if drawings and dissections of this 

 genus have ever before appeared in an illustrated Flora. 



As explained above the work, as first projected, was 

 to be of a semi-popular character, consisting of illustra- 

 tions and short descriptive notes without aiming at the 

 completeness or authority of a Flora. But before it was 

 half-finished a sudden illness drove me to England, and 

 I had there an opportunity of comparing my specimens 

 with those at Kew. I then learnt that a large number of 

 the names, to which we had become accustomed in 

 South India, needed revision : partly because the new 

 material, collected since the Flora of British India was 

 written, has made critical determination more possible 

 and partly because of the different conception of species 

 now held ; while several new species have since been 

 discovered or described. 



More than twenty have been re-named or had older 

 names restored, mostly by being separated from North 

 Indian species with which they had been united. Thus 

 Hypericum wightianum Wall, had been united in the 

 Flora of British India to H. nepalense Choisy, but is quite 

 distinct in having a one-celled ovary with three parietal 

 placentas, while the latter has three distinct cells. 

 Again, Jasminum bignoniaceum Wall, had been reduced 

 to J. humile Linn., a plant of uncertain origin, but obtain- 

 ed in the first instance from Spain or Italy, and different 

 both in leaf and flower. An interesting find was that 

 two species of Dichrocephala, D. latifolia DC. and D. 

 chrysanthemifolia DC. are one and the same, a plant 

 having been found with the characters of the former on 

 its lower, and of the latter on its upper branches. But 

 the chief discovery of this kind was in connection with 



