with short stalks and broad concave limbs. Corolla of tbree lo b e s; imier described by D e CandoUe as toothed, but
qmte entire in my specimens, galeate; lateral lobes obovate. OafsuU liiiear-clavale, truncate, roimded or apiculate
at the apex. Seeds glabrous except for the long silky beard, linear-ohlong; testa green.
2 . Comesperma retusmn (Lab. N o v . H o ll, ii, 2 2 . t . 1 6 0 ) ; erectum, robnstum, ramosuroy foliis
coriaceis obovatis linear i-oblongisve obtusis acutis mucronatisve ( f - | - nnc. lon g .)j racemis terminalibus con-
tractis, capsulis obcordato-spatlmlatis in stipitem elongatnm contractis apice late retusis subbilobis cum
apiculo, sem üiibus carúncula n u lla .— A C F ro d r . I. c . ; S te e tz, P la n t. P re ise , ii. 2 6 6 . [Ounn, 1 7 0 .)
1-I.iB. Abundant, e specially in th e northern parts o f the Islan d , from th e sea to 3 6 0 0 feet elev. on the
We stern Mountains, in a m o ist lig h t soil, Ounn, e tc.— (PI. N o v .) {v. v.)
D i s t b i b . N ew Son th Wale s and South-eastern Australia.
A very handsome-fiowered shrub, 1 -3 feet high, branched from the base.— ® » » and h a n c im stont, leafy, with
pale yellow-brown bark. Leaves very thick, x - 1 inch long, obovate, linear or oblong, blunt, acute or roimded.
Flowers purplish, in terminal subumbellate panicles. Fedmeles short, with linear-oblong bracts. Outer sepah
broadly ob lon g ; two inner large, broadly effiptic, veiy concave, not suddenly contracted at the base. Inner lobe
o f corolla deeply concave and galeate, entire, lateral narrow. Capsules i inch long, upper part broadly obovate or
spathidate, retnse and almost two-lobed at the apex, with a central apiculus, contracted below into a narrow
peduncle. Seeds villous as well as being clothed with long silky descending hairs.
3. Comesperma ericinum (DC. P rodr. i. 3 3 4 ) ; frnticosnm, erectum, gracde, glaberrimum v.
obsonre pubescens, foliis linearibus lineari-oblongisve raiins e lliptico-oblongis obtusis acutis apicnlatisve
marginibns recurvis revolntisve, racemis terminalibus elongatis, capsulis apice trnncato trilobo v . late retusis
medio apiculatis.— P la n t. P r e iss. ii. 2 9 7 , 0 . liu a iiifo lium , A . Ounn. f l i . P . MUller. [Gunn, 6 4 7 .)
Var. a ; foliis sparsis an gu ste linearibus mucronatis.— C. ericinum, B C . I. c . ; S te e tz, I. e. 2 9 7 .
Var. ß . a e u tifo lia ; foliis sparsis oblongis linearibnsve mucronatis.— 0 . acutifolium, S te e tz, I. c. 2 9 6 .
Var. y . la tifo lia j fohis lanceolatis oblongo-lauceolatisve aeutis.— C. latifohum, S te e tz, I. c.
H.1 B. N o r th coast, near th e sea, and islan d s o f B a ss’ Straits, on sandy soil, Gunn.— (FI. N o v .)
D i s t b i b . N ew South Wales a n d South-eastern Austraha.
A slender, erect, twiggy bush, 3 feet high, very variable in habit, foKage, in d size o f dowers.—t o a e s Hnear-
elKptical or oblong, acute, obtnse, or apiculate; i - | inch long, the shortest proportionably broader. A e em e s terminal,
elongated, few- or many-ñowered. Flowers pale purple. S ra c ts very deciduous. Sepals th re e; outer broadly-
ovate; inner very concave, oblong, suddenly contracted to a crmeate base. Inner lobe ot coroUa entire; lateral
linear-oblong, with an obscure auricle on the side next the inner lobe. CapsuU simdar to tbat o f C. retusa, bnt
shorter. Seeds with a strophiolus and very long silky hairs; testa also hairy.— An exceedingly variable plant,
probably common aU over extratropical Australia. Gnnn sends two states : one larger and taller, with larger leaves
and larger flowers ; the other (growing in more exposed situations) much smaller, ivith broader leaves, and smader,
darker-coloured flowers. The flowers and fruit of these states vaiy much, bnt the same variations occur in each.
Midler’s C. linurUfoliurn, A.C., is the same plant, with a sBght pubescence ; Steetz’s amtifolium and Utijolium are
sent by Gunn as varieties.
4 . C o m e s p e rm a C a ly m e g a (Lab. 1. o. p . 2 3 . t. 1 6 2 ) ; glaberrimum, herbaceum, erectum, e basi
ramosum, ramis strietis virgatis simplicibus, foliis radicahbus (dum adsunt) majoribus elliptico-lanceolatis,
caulinis sparsis lineari-laiiceolatis omnibus acuminatis coriaceis, racemis ramos termiiiantibus e longatis m ultifloris,
floribus breve pedicellatis, sepalis 3 exterioribus magnis oblongis obtusis interioribus 2 late ovato-
rotundatis brevioribus, petalis lobo intermedio integro lateralibus longioribus linearibus acuminatlb, semi-
iiibus v illosis e t lon g e lanatis, carúncula im\lo..— I)C . F r o ir . I. e. ¡ Hook. Comp. Bol. Mag. i. 2 7 4 ; S te e tz
n i - i o . . . . ' - ;; o o » isocalyx, Spreng. S y s t. Veg. iii, 1 7 2 ( f t i . S in. R ia n t. P r e is s . ii. « te e tz ). [Gunn, 7 3 5 .)
IÎAB. Common on sandy flats along the north shores o f the Island, and in the islands o f Bass’ Straits,
Gunu.— (FI. N o v . D e c .)
D i s t r ib . South-eastern Australia.
A very distinct and pretty little species, the only one in Tasmania, except C. volubile; ivith bright blue
flowers, which sometimes vary to purple and white.— Rgot woody, sending up many ascending and erect slender
branches, a span to a foot high, terminated hy many-flowered conical racemes. Leaves scattered on the stems, | - 1
inch long, linear or lanceolate, acuminate, the radical broader aud larger. Racemes 2 - 4 iuehes long. Flotcers
smaller than in any of the foregoing species, on shorter pedicels. Sepals three ; outer larger, oblong, membranous,
tipped with blue ; three inner membranous, bright blue, rarely white or purple, veined, rounded, coucave. Corolla
cleft to the middle ; inner lobe galeate, entke ; lateral elongated, acuminate, narrow, membranous, blue. Capmle
small, -i- inch long, less conspicuously truncate than ki C. Calymega and C. retusa, ohcordate at the apex, with
rounded lobes and a central apiculus. Seeds broivn, villous, and also bearded with hairs that are more flexuous
and cottony, or woolly, than in the preceding species, in which they are silky.
N a t . O r d . IX. TREMANDREÆ.
A smaE N atural Order, comprising three genera and twenty-five to thirty species, aU confined to the
Australian Continent and Tasmania, and most o f them b e lon gin g to th e g en u s Te tratheca. Several abound
in Tasmania and Eastern extratropical Australia, bu t th e majority are found in South-western Australia;
one from Endeavour River is th e only tropical species known. They frequent sandy heaths or dry places,
aud imitate th e E riceæ o f Europe and the Cape o f Good H op e in a very marked manner.
T he affinities o f Tremandreæ are with P o lyg a leæ , as th e structure o f their flowers and fru it indicates ;
and th a t this is their nearest ally was first indicated b y Brown. Planchón and S teetz however consider
that they are more nearly allied to B y ttn e ria c eoe , through L a siope talum and it s allies, with which th ey have
many p oints o f structure in common.
Gen. 1. T E TR A TH EC A , Sm.
a 4>-5. P e ta la 4 - 5 , æstivatione induplicata. S tam in a 8 - 1 0 , subæqualia. Antheroe cylindricæ,
2 -4 -lo cu la r e s, poro terminali dehiscentes. Ovarium 2-(rarius 4 -)loculare, lo cu lis 1 -3 -o v u la tis. Capsula
obovata, 2-locularis (rarius 4 -locular is), loculicide 2-vaIvis. Semina 1 - 2 , chalaza carúncula fungosa
sæpe torta aucta. — FruticuR g r a c ile s ; foliis a lte rn is v e rtic illa tis v e , e s tip u la tis ; pedunculis g ra cilibu s,
axiUaribus.
Tlie Tasmanian species o f this genus have long been a subject o f detailed study by Mr. Gimn and myself, and
the results are very unsatisfactory. I examined them first in 1838, and concluded that there were but two permanently
distinct ones, T. ciliata and another more common form, which I divided into four varieties (Joum. B ot. p. 408).
When in Tasmama Mr. Gunn and I v eiy carefully studied living and diied specimens, and came to the conclusion that
four kinds were distinguishable, o f which however three could not be chai-acterized hy any trenchant characters, though
they could generally be recognized when groiviug. Since my retmm from Tasmania I have twice examined the genus,
taking the Tasmanian species both by themselves and iu ooujunction ivith the Australian, and find no reason to alter mv
original opinion, that there arc no absolute characters by whicli to distinguish, permanently, more than two species ; but
that, in accordance ivith Mr. Gimn’s and my oivn later opinion, the difl’erences between the prevalent forms o f one
species admitted o f its being divided into four varieties. Since the period alluded to, Mr. Gunn has detected a fifth and
sixth form or species, T.procumhensaaà. T. Gunnii. He remarks that the extreme forms retain tliek character iiith considerable
constancy throughout the Island. This is, in short, one of those cases that so frequently occur in systematic
VOL. I . K