HANDBOOK OF THE BROMELIACEJE. 25 



the text, Professor Morren's water-colour drawings have heen an 

 important aid; " described from Prof. Morren's drawing " occurs 

 repeatedly at the end of descriptions, not only of Morren's own 

 species, but also of new ones which Mr. Baker has been able to 

 found on the sketches. In the genus Cryptanthus, containing 

 twelve species, five are thus described, one C. Makoyanus for the 

 first time. " M.D." in brackets after the name signifies, as we are 

 told at the bottom of page 2, that there is an original drawing of 

 the plant in the Morren Collection at Kew. The end of the Preface 

 would have been a better place for such a note. Apropos of figures, 

 Mr. Baker quotes these at the commencement of each description, 

 and it is a great help to have his opinion as to what some of them 

 are meant to represent. He has run down nearly all the plates of 

 Vellozo's 'Flora Fluminensis " ! Except for occasional misprints, 

 these references are mostly correct. Horticulturists will welcome 

 the addition of the date of introduction and first-flowering of culti- 

 vated species. 



The Order is divided into the same three great Tribes as in the 

 1 Genera Plantarum,' viz. : — 



I. BromelieeB, characterised by an inferior ovary and berried 

 fruit, and leaves nearly always spine-margined. 



II. Pitcaimiece, with capsular fruit and seeds of which the 

 funiculus does not break up into fine threads, and 



III. Tillandsiece, where the capsular fruit is always entirely 

 superior, the funiculus breaks up into threads, and the leaves are 

 always without marginal prickles. 



We mention this, because Wittmack, who has written the 

 Bromelmcem for Engler and Prantl's ' Pfianzenfamilien,' separates 

 Puya, Encholirion , Dyckia, and Hechtia as a distinct tribe, Puyece, 

 distinguished from Pitcaimiece (Brocchinia and Pitcairnia) by 

 its entirely superior ovary, and from Tillandsiece, by its toothed 

 leaves. The new tribe does not approach in distinctness to the 

 three older ones, which are so well marked and natural. Pit- 

 caimiecB, as Tribe II, of the present arrangement, gives us a perfect 

 gradation between the epigynous Bromelice and the perfectly hypo- 

 gynous Tillandsiea. The Key to the Genera, on page x. of the ' Hand- 

 book,' shows this at a glance. First Brocchinia and Andre's new 

 genus Bakeria, with the capsule free only near the tip; then Pit- 

 cairnia, with the capsule free except near the base ; and finally Puya, 

 Cottendorfia, Dyckia, and Hechtia, all with a capsule entirely superior. 

 Some of the Pitcairnias, forming the section Puyopsis exactly 

 resemble Puya in leaf and habit, being only distinguishable by the 

 loculicidal dehiscence of its capsule, a character which separates it 

 quite as much from Dyckia and Hechtia. 



The subdivision of the first tribe, Bromelice, is much the same 

 as that followed in the ' Genera,' where we find three divisions 

 depending on the freedom or complete or partial coherence of the 

 sepals and petals. Mr. Baker makes two divisions ; in the first, 

 comprising Karatas, Greigia, and Distiacanihus, the petals are united 

 into a distinct tube, which is usually as long as the calyx-limb. 

 In the ' Genera' the first group includes Streptocalyx and Bromelia, 



