June 16, 1916] 



SCIENCE 



859 



small to account for the great changes which 

 occur in living tissue. 



It might be supposed that the reason that no 

 change in resistance occurs in dead tissue is 

 because the hardening and softening do not 

 proceed as far as in living plants, but this is 

 not the cas^. Moreover, it is found that the 

 increase of viscosity in NaCl is accompanied 

 by absorption of water, while the decrease of 

 viscosity in CaCl, is accompanied by loss of 

 water, and these processes take place in the 

 same way in living and dead tissue. 



It would seem that these and other impor- 

 tant objections must be removed before we can 

 accept the idea that changes in permeability 

 are determined by changes in viscosity.* 



w. j. v. osterhout 



Laboratory op Plant Physiology, 

 Harvard University 



pollen sterility in relation to 



CROSSING 



In view of the recent revival of the old idea 

 that pollen sterility is a universal and safe 

 criterion of hybridity in plants^ we found it 

 of interest recently to examine the pollen of 

 some California plants with this idea in mind. 



The first species examined, Trillium sessile 

 var. giganteum, perhaps better regarded as T. 

 giganteum, a separate species from the T. 

 sessile of the eastern states, is found in quan- 

 tity in Strawberry Canyon, Berkeley, where it 

 is now in full bloom. It is already known that 

 this species shows a remarkable degree of 

 variability, especially in the color and width 

 of the petals. In color the petals vary from 

 dark purple through pinks to nearly white, 

 and also through yellows to nearly pure green. 

 One of us is making a detailed study of these 

 variations. The former color series, com- 

 bined with the width series, is found on one 

 hillside in Strawberry Canyon, the greenish 

 and yellowish series occurring across the bay 

 in Marin County. No other Trillium occurs 



4 It would appear that the term viscosity is 

 loosely applied to a variety of phenomena which 

 may be produced in different ways. 



1 Jeffrey, E. C, 1915, ' ' Some Fundamental 

 Morphological Objections to the Mutation Theory 

 of DeYries, ' ' Amer. Nat., 49 : 5-21, Pigs. 7. 



in this canyon, but a variety of T. ovatum 

 occurs along with T. giganteum in various 

 parts of Marin County. The two forms are 

 not closely related, however, and it is ex- 

 tremely doubtful if they ever cross. In 

 Strawberry Canyon at any rate there is no pos- 

 sibility of T. giganteum crossing with any 

 other species, ^et some plants collected here 

 show a considerable amount of sterile pollen. 



In all the pollen examinations the grains 

 were only considered " bad " when they were 

 obviously shrivelled or greatly undersized, so 

 that the amount of non-viable pollen would 

 doubtless be considerably larger than the per- 

 centage recorded here as bad. The highest 

 amount of bad pollen recorded from any nor- 

 mal plant of T. giganteum, from Strawberry 

 Canyon was 18.2 per cent., and the lowest 

 3.2 per cent. In another plant having certain 

 abnormalities of the flower the percentage was 

 as low as 1.5 per cent. In five plants from 

 Camp Taylor, Marin County, where the spe- 

 cies grows in company with T. ovatum, the 

 percentages of bad pollen were respectively 

 Y.3, 5.6, 3.9, 3.2, 2.3. Thus the amount of de- 

 fective pollen is not high in any of the plants 

 examined, with one exception, though the 

 pollen grains are never all perfect. 



The form of T. ovatum occurring in Marin 

 County is remarkably uniform, in contrast 

 with the variable T. giganteum. The pollen 

 from seven plants of T. ovatum was examined, 

 and they were found to have respectively Y.3, 

 Y, 5.3, 4.5, 4.2, 3.9 and 3.9 per cent, bad pollen 

 grains. Thus a species which is very invari- 

 able in this locality and which we can be quite 

 certain does not cross with T. giganteum, 

 nevertheless produces regularly a certain per- 

 centage of shrivelled and misshapen grains. 



Still more conclusive evidence regarding 

 the occurrence of considerable quantities of 

 bad pollen in the absence of crossing was fur- 

 nished by Scoliopus. This remarkably iso- 

 lated genus of the Liliacese contains only two 

 species, 8. Bigelovii, which is confined to Cali- 

 fornia from Santa Cruz to Humboldt Coumty, 

 and 8. Hallii, which occurs in western Oregon. 

 In plants of 8. Bigelovii collected in Marin 



