992 CLASSIFICATION OF PLANTS. 



sive, as Fungi and Algo}, each of which is divisible into tribes more distinct on the whole 

 (whether strictly limited or not) than are most allied Orders of Phsanogams from one 

 another. In other words, the Orders of Cryptogams and Phsenogams have very different 

 absolute and relative values. 



MoNOCOTTLBDONS. — In the classification of these, the three following structural characters 

 have been regarded as of primary importance. 1. The relations of the floral envelopes to 

 one another and to the ovary, by Jussieu (followed by Fries), who divided them into 

 epigynous, hypogynous and perigynous, in conformity with his division of Dicotyledons 

 (see p. 164). 2. The presence or absence of floral envelopes, and their nature ; whence 

 originated the Petaloid and Glumaceous divisions of Lindley (subsequently modified in his 

 'Nixus'). 3. The presence or absence and nature of the albumen, as proposed by 

 Brongniart. An important subsidiary character, derived from the position of the embryo, 

 whether enclosed in or external to the albumen, is also of great importance. 



In the classification I have proposed, the greatest importance is given to Jussieu's 

 character of the adhesion or not of the perianth to the ovary ; this appearing to afford the 

 means of subdividing the Monocotyledons into two series better than does any other single 

 character, or combination of characters. The exceptions are few, and the typical Orders 

 with inferior perianth are, on the> whole, more closely allied to one another than to any in 

 the series with an inferior perianth ; but the character of perigynism and hypogynism does 

 not answer for primary divisions. 



This disposition of the Monocotyledons appears to follow a simple series, descending 

 from, the Orders most complex in structure (Orohidece, &o.) to those most simple 

 (^Gh-aminece) ; the principal exceptions are the Incomplete- flowered Orders, Aponogetce and 

 Naiadece, which are obviously reduced Alismaaeoe, and Lemnacece and Typhaeece, which are 

 obviously closely allied to Aroidece. 



Angiospermous DiooirLEDONS. — A practicable linear classification of the Orders of 

 this division, which shall also be natural, has not hitherto been proposed, nor is it likely 

 that it is attainable. In the first place, it is not so obvious in what a high type of 

 structure consists ; and in the next place, so many of the Orders contain both high and low 

 types, and in such various proportions, that it is impossible to establish a table of prece- 

 dence for them. De CandoUe suggested that plants whose flowers had the greatest number 

 of distinct and separate organs should head the series, namely, those with hermaphrodite 

 polypetalous flowers, having apocarpous ovaries ; and that unisexual plants with incomplete 

 perianths or none should terminate it. 



Succeeding authorities have considered departure from the primitive condition of the 

 floral whorls (regarded as foliar organs) as evidence of high type, and consequently that 

 plants with consolidated floral organs should head the series ; placing Monopetalte before 

 Polypetalss, and Monopetalse with the calyx-tube and ovary consolidated (epigynous Mono- 

 petals) before those in which the ovary is free. It has further been held that a flower in 

 which many such consolidated organs are subsidiary to the perfecting of one seed is of a 

 higher type than one in which as many organs are subsidiary to the perfecting of many 

 seeds, and that Oompositce, by satisfying the first of these conditions more nearly than any 

 other Order, should hence head the series. 



Whatever the difference of opinion as to the value of cohesion of the petals, that is, of 

 monopetalism in contradistinction to polypetahsm, as evidence of high or low type of 

 structure, there is none as to its value as an index of afiinity ; the Monopetalous Orders 

 being indisputably more nearly related on the whole to one another than they are to the 

 Polypetalous. It is very different with the Orders of Jussieu's and De Candolle's third 



