68 BOTANICAL GAZETTE. 



as the rock itself; only when accidentally it has taken root in a coarse gravelly debris 

 can it be successfully removed witli its terminal lihroiis roots. The main stem, often 

 attaining a thickness (*f an inch or more, divides horizontally into radiating gnarled 

 branches, which send otl' from the axils of the primary leaves numerous leafy rosettes 

 by which the growth of the plant is maintained, and from the axis of svhich the slender 

 bracted tlower stems arise. These stems trom 3 to « inches high are crowned for about 

 one-third their length with a dense cyliadrical spike of small white tlowers, with ex- 

 serted style and stamens, set in a light green tomentose calyx. In thrifty specimens these 

 flow^ering spikes are occasionally liranched. The prf)strate growing shoots whenever 

 they light upon a suitable crevice strike root and thus iielp to maintain the growth of 

 the parent plant ; where these favorable conditions are wanting, the whole plant, often 

 covering several square feet, can be lifted up. showing on the under surface the remains 

 of dead leaves, in the decaying mould of which moisture is retained to keep up the 

 fresh verdure of the young twigs.. It seems to llourish best on the north side of ex- 

 posed limestone rocks occupying a considerable range of altitude above the lower 

 foothills, choosing by preference the nigged slopes of moderately steep canons. Its 

 flowering period is earliest on the higher elevations, continuing from the latter part of 

 August to the tirst of October. The peculiar adaptation of this plant for ornamental 

 rock work can be appreciated by those wiio have once seen it in its native haunts, and 

 it is hoped that from plants and seeds somewhat copiously collected it may eventually 

 tind a much larger number of admirer.^ in gardens devoted to this charming class of 

 horticultural adornments." 



A Stkange Collinsia. — During tlie season 1 received a tlower for determina- 

 tion, whicli at first glance 1 jtronounced to be (JnlUnsia centa, Nutt. (.)n closer 

 examination however I found the sj^ecimens I had presented two marked variations 

 from the typical jilant. The whole were invariably 8-flowered instead of ''about six." 

 Then the corolla instead of being only "twice the length of the cali/x" was in every in- 

 stance fully three times its length. On searching the loi-ality in which they had been 

 found, I discovered that an are;i of al)out one-half an acre was thickly covered with 

 the plant, and that fully a third of the specimens presented these variations. Hereto- 

 fore all my observations of Collinsia have led me to believe it to l)e a ctmstant species. 

 The locality in which these irregular Collinsias occur was a moist ridge slightly above the 

 level of the surrounding swamp. The soil, the rich black loam characteristic of our 

 Indiana prairies. I regret exceedingly that having no facilities for i)reserving them I 

 secured no specimens. M. S. C. 



In the italicised characteristics of lludbecLia hirtti, L., 1 find, upper leaves s/'sxile. I 

 hav'e specimens in which the upper leaves are distinctly petioled. The petiole Ijroad- 

 margined. It may be possible that the specimens are in an imperfect state, although 

 it seems to me, every other characteristic being perfect, that there can be no doubt that 

 this is a variation from the typical plant. I noted it only on account of its differing 

 from the description in (Tray's Manual, and hecau.se I deemed these slight departures 

 fnmi the typical form of interest t(i botanists. M. S. C. 



A.II cominKiilixitiotif^ s/iDtf/d f)e addres,sed to 



John M. I'oultrr, Haytovev, liul. 

 M. S. Coulter, Loo'ansport, litd. 

 Terms: — Subscription $1.00 a year. Single y umbers 10 cents. 



