Retz. This limitation is substantially in accordance with the limitation of the same two Acalyphas by Meyer in 1843; the plant identified with A. glabrata, Thunb. is in reality that species while the one identified as A. betulina is not the plant so named by Retz but is a variety of A. glabrata. The last Acalypha in this list, Zeyher 3840, identified correctly as A. discolor, E. Mey., is, therefore, as we have seen, A. capensis (= A. decumbens, Thunb.).

This Zeyherian list, briefly summarised, therefore adds to the genus two additional South African species neither of which is provided with a name.

The next contribution to our knowledge of the South African species of Acalypha was published by Sonder in Linnaea, vol. xxiii. (1850), pp. 115-117. Through the kindness of Professor Lindman and Dr. Dahlstedt we have been enabled, in considering this enumeration, to examine the actual specimens with which Sonder dealt. We may discuss the six species he accounts for seriatim.

The first species recorded by Sonder is "A peduncularis, E. Mey.! Meisn. syn. A. crassa, Buching!" for which he cites no specimen but to which he adds a variety "B. glabrata, Sond." based on a specimen collected by Zeyher on the Macalisberg Range. The only specimen in his herbarium on which Sonder has written the name "A. peduncularis, E. Mey." is one of Zeyher 3838, from Howison's Poort, Albany Div., which is exactly like the plant collected in the Assegai Bush, Albany Div. by Drège, and obtained by the same collector on two occasions on the Zuureberg Range, that forms the basis of A. peduncularis, E. Mey. It so happens that Sonder's statement that this plant is really conspecific with the Natal one described by Meisner under the same name, is accurate. But an examination of his specimens shows that the accuracy of Sonder in this regard is purely accidental, for he did not have at his command any example of any portion of Krauss, 377. The specimen which led Sonder to this conclusion is one collected by Gueinzius at Port Natal which was written up in Sonder's collection as "Acalypha peduncularis, E. Mey. ?-crassa, Buch.!" Above this legend has been added subsequently the name "A. punctata, Meisn. ?," a justifiable afterthought because this plant of Gueinzius does not agree either with the part of Krauss 377 which Meisner described as A. peduncularis, or with the part of Krauss 377 which Buchinger named A. crassa. It is, however, identical with the remaining portion of Krauss 377, which is the type of A. punctata, Meisn.

Nor has Sonder been really more fortunate in his treatment of his own variety 'glabrata.' It is true that the description given by Sonder applies only to one plant, and that this plant is a specimen of Zeyher 1521, from the Macalisberg Range, on which Sonder has written the name A. peduncularis, B. glabrata. But in his own herbarium Sonder wrote up a specimen of Zeyher 3839 also as A. peduncularis, ẞ glabrata; to this second specimen, which was collected on Van Staadensberg, Uitenhage Div., Sonder's description of var. glabrata is quite inapplicable.

Sonder's second species is one which he describes as new and is based on Gueinzius 171, from Natal and on Zeyher 1518, from the Macalisberg Range. This species, A. angustata, Sond., is perfectly

valid and the only modification that is required in his view consists of the inclusion of the proposed variety within the type.

Sonder's third species is A. glandulifolia, Buching., which we have already seen to be a valid species. Here again Sonder has suggested the recognition of two distinct forms; the advent of additional material indicates that this differentiation is not required.

The fourth species enumerated by Sonder is the one first issued by Drège in 1843 as A. languida, E. Mey., but first described. by Hochstetter in 1845 as A. petiolaris. In giving preference to the synonym which, though the older, is merely a naked name, Sonder acted unfortunately.

The fifth species is that which Sonder has termed A. betulina, Retz. The species is now for the first time dealt with intelligibly and the true relationship of the two allied forms, whose existence had already been indicated by E. Meyer and by Ecklon and Zeyher, is more clearly defined. But the name employed is unfortunate; the species is not A. betulina, Retz, but is A. glabrata, Thunb. Fuller knowledge, moreover, indicates that there is no real necessity for the recognition of var. latifolia as apart from the type.

The sixth species, in spite of the doubt to which Sonder testifies, is really an Acalypha and is, as Sonder indicates, a distinct and valid species. But the use, in designating this new species, of the epithet which Sonder was aware Hochstetter had already applied to another Acalypha is singularly unfortunate. Our present conventions, which render incumbent the use of the name "A. petiolaris" in connection with the plant described as such by Hochstetter in 1845, prevent us from employing it to designate the plant so described by Sonder in

1850.

In Linnaea, vol. xxv. (1852), at p. 587, Scheele based on Drège, 8242, a species of Acalypha from South Africa, A. lamiifolia, Scheele. The plant on which this species was based is, however, identical with that to which Buchinger bad already given the name A. Kraussiana and is hardly distinguishable from Urtica capensis, Linn. f. Tragia villosa, Thunb., of which A. decumbens, Thunb. is at best only a variety.

=

Ten years later Baillon published in Adansonia, vol. iii. (1862), pp. 156-158 a resume of the South African species of Acalypha, based partly on specimens, partly on the literature which has been passed under review.

1. Acalypha peduncularis, Baill. is identical with A. peduncularis, E. Mey., the specimens of Masson and of Zeyher 3838 which are cited agreeing precisely with those of Drège on which the species was based. Baillon, however, had no opportunity of seeing any specimen of Krauss 377 and his erroneous reduction of A. crassa, Buching. to A. peduncularis is adopted from Sonder.

2. Acalypha Zeyheri, Baill., based on Zeyher 3839 and on a specimen of doubtful provenance bearing the number 301, supplies the earliest description of a valid species.

3. Acalypha caperonioides, Baill., based on Zeyher 1521 is again a valid species. In this instance Baillon has failed to note that the same plant is the type of A. peduncularis, ẞ glabrata, Sond.

4. Acalypha angustata, Baill. is Sonder's species of this name.

5. Acalypha glandulifolia, Baill. is Buchinger's species of this

name.

6. Acalypha languida, Baill. is E. Meyer's species of this name treated as it was treated by Sonder.

7. Acalypha betulina, Baill., to which Baillon only doubtfully refers Zeyher 1517a, the only specimen seen by him, is on this account A. glabrata, Thunb. proper.

8. Acalypha discolor, Baill. is A. discolor, E. Mey. treated as it was by Krauss and Hochstetter in 1845. To this Baillon has added. a variety ẞ major, Baill. which is, though Baillon was unaware of the fact, the same as A. cordata, Thunb.

9. Acalypha lamiifolia, Baill. is A. lamiifolia, Scheele, the identity of which with A. Kraussiana, Buching. Baillon has failed to notice. It is also almost identical with Urtica capensis, Linn. f. = Tragia villosa, Thunb., of which A. decumbens, Thunb., is only a variety.

10. Acalypha brachiata, Baill. is E. Meyer's species of this name. 11. Acalypha Eckloni, Baill., based on a gathering issued by E. Meyer as A. cordata?, is identical with the preceding species. As the only description so far published was that here supplied by Baillon his name A. Eckloni supplants the earlier name A. brachiata.

Briefly summarised the resumé of Baillon of 1863 makes us aware of the existence of nine species, viz:-A. glabrata and A. capensis already known to Thunberg in 1800; A. petiolaris, A. Eckloni and A. peduncularis already known to E. Meyer (as A. languida, A. brachiata and A. peduncularis respectively) in 1843; A. glandulifolia already known to Krauss in 1845; A. angustata already known to Sonder in 1850; finally A. Zeyheri and A. caperonioides the existence of which had been indicated by Ecklon and Zeyher in 1847, now for the first time properly named and described. Baillon's resumé fails to account for A. punctata, Meisn. published by Krauss, or to observe that A. lamiifolia, Scheele is hardly more than a repetition of A. Kraussiana, Buching. also published by Krauss.

This resumé was followed in 1865 and 1866 by the preliminary and the finished monographs of the genus by Müller published in Linnaea, vol. xxxiv. (1865), pp. 1-54 and in De Candolle's Prodromus, vol. xv., pars ii. (1866), pp. 799-889 respectively. These two accounts we may conveniently consider together.

In the earlier account, Müller gives ten species as coming from South Africa. These are:-10, Sonderiana (Linnaea, vol. xxxiv., p. 9); 87, peduncularis (p. 28); 88, Zeyheri (p. 29); 89, petiolaris (p. 29); 90, languida (p. 29); 91, tenuis (p. 30); 92, Eckloni (p. 30); 98, glabrata (p. 36); 118, discolor (p. 38); 119, Kraussiana (p. 39).

In the fuller account of the following year, Müller enumerates eleven South African forms, adding two new species to the 1865 list and at the same time reducing two of those in the earlier list, discolor and Kraussiana, to the position of varieties of one species, decumbens. The 1866 list is as follows:-10, Sonderiana (DC. Prodr. xv., ii., p. 804); 59, grandidentata (p. 823); 116, peduncularis (p. 846); 117, Zeyheri (p. 847); 118, petiolaris (p. 847); 119, languida (p. 848); 120, tenuis (p. 848); 121, patens (p. 848); 122,

Eckloni (p. 849); 139, glabrata (p. 857), and 156, decumbens (p. 864). In addition to the foregoing, five of the species described by Müller in 1866, which were not at that time known to occur south of the Tropic, have since that date been gathered in South Africa. These are:-85, ornata (p. 833); 115, senensis (p. 845), from which 114, zambesica, is not distinguishable; 165, indica (p. 868); 175, ciliata (p. 873), and 183, segetalis (p. 877).

No remark is called for in the case of any of these five species, nor is any remark called for in the case of 10, Sonderiana, Müll. Arg. (1865), which is A. petiolaris, Sond. (1850), not of Hochst. (1845), and is a valid species.

Little remark is required in the case of 139, glabrata, Thunb. (1800), which is the plant described in Schultes' edition of Thunbberg's Flora Capensis (1823), and on this account is a valid species notwithstanding the fact that it is not based on the specimen cited in that work, and is not the plant named A. glabrata in Thunberg's own herbarium. In his acceptance of the variety latifolia, Müller has merely followed Sonder. Another valid species is 156, decumbens, Thunb. (1800), now A. capensis, which is identical with A. discolor, E. Mey. ex Meisn. (1845), and includes Urtica capensis, Linn. f. (1781), Tragia villosa, Thunb. (1794), A. cordata, Thunb. (1800), A. Kraussiana, Buching. (1845), A. lamiifolia, Scheele (1852), and A. discolor, ẞ major, Baill. (1862). But 59, grandidentata, Müll. Arg. (1866), is identical with Tragia villosa, Thunb., and therefore is not a valid species. Owing to Müller having failed to observe that Urtica capensis, Linn. f. (1781) and Tragia villosa, Thunb. (1794) are the same plant, he has omitted to employ for the species as a whole the name "Acalypha capensis.' Another valid species is 122, Eckloni, Baill. (1863), which includes A. brachiata, E. Mey. In 1862 Baillon had himself actually accepted this name which Meyer had proposed in 1843. But in so doing, Baillon, like Meyer himself, failed to prepare a description; as a consequence, the conventions of nomenclature insist that the name A. Eckloni, though the accident of a misconception, shall be used for Meyer's plant.

Another quite valid species is 118, petiolaris, Hochst. (1845), the name used for which similarly supplants the name A. languida proposed by Meyer two years earlier. In this instance, however, Müller has made an effort to maintain A. languida, E. Mey., as a species. His justification for this is the belief that a specimen in Herb. Berlin, which Meyer has written up as A. languida, differs specifically from the other specimens so named, and therefore from A. petiolaris, Hochst. This view cannot be sustained. There is no specimen in Herb. Berlin named A. languida by Meyer which differs in any important feature from A. petiolaris, Hochst. Therefore 119, languida, Müll. Arg. (1865) is not a valid species. The same remark is called for in the case of 120, tenuis, Müll. Arg. (1865), to which Müller has attributed two varieties which do not differ from each or from A. petiolaris, Hochst., by any tangible

character.

In the case of 117, Zeyheri, Müller has deviated very considerably from the treatment accorded to that species by Baillon when he founded it in 1863. Baillon's original types were Zeyher,

3839, and a specimen of doubtful provenance bearing the number 301, which Müller has been able to assign to Krebs. This latter specimen Müller transfers to A. peduncularis, var. psilogyne, Müll. Arg. But Müller's modification of A. Zeyheri by no means ends here, for he treats A. Zeyheri, Baill. (1863) as merely a variety, var. glabrata, of an enlarged A. Zeyheri, Müll. Arg. (1864), the other variety of which, var. pubescens, is in intention identical with the original A. peduncularis, E. Mey. (1843). There is no clue in the earlier account given by Müller in Linnaea, vol. xxxiv., to the idea underlying this arrangement. Apparently as an afterthought, a character is added in the account given in the Prodromus which might have served to differentiate A. Zeyheri, Müll. Arg., hardly of Baill. from A. peduncularis, Müll. Arg., not of E. Mey and hardly of Meisn., had it been constant. This character is that in A. Zeyheri, as widened by Müller, the flowers are monoecious, whereas in A. peduncularis they are dioecious. It is true that in some, but by no means all, of the specimens actually included by Müller in his A. Zeyheri, var. pubescens, the flowers are monoecious. But there is not a single example of Zeyher, 3839, which is all that Müller has left in A. Zeyheri, var. glabrata, in which the flowers are monoecious. In his choice of the varietal name glabrata to designate Baillon's original A. Zeyheri, Müller has been singularly unfortunate. The name was selected under the impression that the plant in question is A. peduncularis, var. glabrata, Sond. We know, as Müller clearly also knew, that Sonder had without due consideration, written this name upon his example of Zeyher 3839, the gathering on which A. Zeyheri, Baill., was based. But this does not alter the fact that Sonder's description of A. peduncularis, var. glabrata, does not apply to this plant, or modify the circumstance that the plant on which Sonder's description of A. peduncularis, var glabrata, was based, was treated by Baillon as the basis of a distinct species, A. caperonioides, Baill., which Müller, in 1866, though not in 1865, has referred as a distinct variety to his A. peduncularis.

means

The same entanglement marks Müller's treatment of his A. Zeyheri, var. pubescens. The specimens collected by Masson, Krebs, and Ecklon and Zeyher, which Baillon refers to A. peduncularis, var. genuina, are identical with those collected by Drège, which are referred to A. Zeyheri, var. pubescens. The latter are by no means invariably monoecious, the former by no invariably dioecious, and the idea that Meisner was in error when he identified with Meyer's plant so named a species from Natal which he described for Krauss as A. peduncularis, is without foundation. At the same time A. Zeyheri, Baill., remains a perfectly valid species, but one with which A. peduncularis, var. glabrata, Sond., is not synonymous.

as

If the treatment accorded by Müller to A. Zeyheri, Baill., leaves something to be desired, this is more markedly the case regards the treatment of A. peduncularis. In this species, Müller has recognised as many as seven varieties :-(a) caperonioides ; (B) genuina; (7) psilogyne; (8) crassa; (e) punctata; () glandulifolia; (n) angustata. Of these a, caperonioides is a valid species, A. caperonioides, Baill. (1863), the earliest name for which, A. peduncularis, var. glabrata, Sond. (1850), Müller has transferred, as a