168 BOTANICAL GAZETTE. 



the spirit can be seen, before the cell becomes filled, running back- 

 ward and forward on account of capillary attraction and the decrease 

 in diameter occasioned by the twisting. — C. R. Barnes, il/«rf/.so», 7/ic?. 



Trimorpiiism in Lithospermum cankscens, Leiim. — Having occasion 

 last year to collect and study specimens of this interesting species, 

 I found, in the course of my examinations, a j^lant with llowers dif- 

 fering from the ordinary dimorphous condition. This discovery led 

 to an extended search, the results of which may be interesting. 



L. canescens is common in Michigan, as is also the closely related 

 and more showy L. hirtnm., growing in bunches and blossoming in 

 spring as the former does. The masses of deep orange blossoms 

 contrast linely with the bare, brown hillsides and the light green of 

 the newly leaved over-arching trees. Every gardener knows what a 

 pleasing effect is produced by the massing of color, and I thought, as 

 I gazed on the flaming hill-slopes, that the plant w^as really handsome 

 and worthy a place in our gardens. It would, however, scarcely be 

 worth cultivating, if that were possible, since the grace of the sur- 

 roundings w'ould be wanting. This, of course, adds much to the 

 charm of herborizing. If we could carry the freshness of the fields 

 and the delight of discovery into our gardens along with the plants, 

 we should, I dare say, transplant half the herbs we find in our ram- 

 bles. But, to return to the facts, Brof. Gray says in the Manual that 

 dimorphism sometimes occurs in this genus, and Sir John Lubbock 

 has described a number of Borrageworts in which contrivances for 

 cross-fertilization exist; but I have come across no special mention 

 of L. canescens. 



I observed three principal flower forms. One form, which I will 

 call No. 1, has a short style, about one-third the length of the corolla 

 tube, with stamens set high up on the tube in the throat. Sometimes 

 the anthers of this form are somewhat exserted, but generally 

 slightly included. 



No. 2 has styles varying from considerably included to strongly 

 exserted with stamens near the l)ase of the tube, much below the 

 stigma. It seemed to me I could make out three forms of this No. 2, 

 de])ending on the length of the style : (a) with style much exserted — 

 there were but few of this kind; (b) with style up to but not project- 

 ing from the throat of the corolla — this was the common form; and 

 (c) with style about two-thirds the length of the tube. 



No. 3 has the short style of No. 1 with the short stamens of No. 2, thus 

 briiiging the stamens and pistils close to each other in the bottom of 



