September 3, 1886.] 



SCIEJSrCE. 



221 



shown that Argelander's scale in the ' Diirch- 

 musterung ' was just as close to this as that of any 

 single one of the photometers. 6'='. Coming to 

 accidental errors, Mr. Chandler showed that, 

 from a full discussion of the naked-eye esti- 

 mates of Gould, Sawyer, and himself, the prob- 

 able error of a single estimate was a little over 

 ±.06 of a magnitude when the stars wei-e at con- 

 siderable distances from each other, and about 

 ± .05 of a magnitude when near ; while the probable 

 error of a single measure in the ' Harvard photom- 

 etry' was ±.17 of a magnitude, and in the 

 ' Uranometria Oxoniensis ' about ±.10 of a magni- 

 tude, thus showing that the eye-estimates were 

 from two to three times as accurate as the jjhoto- 

 metric. 7*^. Discussing the cause of the large 

 residuals in the ' Harvard photometry,' Mr. Chan- 

 dler showed the strong probability of wrong iden- 

 tification of stars in many cases, citing one case 

 where no bright star existed in or near the 

 place called for by the observing-! ist, on accou.nt 

 of a misprint in the ' Durchmusterung,' and yet 

 some neighboring star was observed on several 

 nights for it. 8°. Also the method of applying a 

 correction for the mean value of the atmospheric 

 absorption was very questionable, since over- 

 whelming evidence pointed to an enormous differ- 

 ence in this absorption from night to night. 9°. 

 The author pointed out that we must obtain better 

 results from photometers if we ever expect to use 

 their results for the detection or measurement of 

 variable stars, since several variable 5 have been 

 detected, and their periods and light-curves well 

 determined, by careful eye -estimates, whose whole 

 range of brightness is no greater, or even less than, 

 the range of error in the photometric observations 

 upon a single star with the meridian photometer. 



In a discussion of a paper by Mr. Barnard upon 

 ' Telescopic observations of meteor-trains,' Pro- 

 fessor Newton pointed out that the study of their 

 drift was the only method we have of studying the 

 upper cui-rents of our atmosphere, except such 

 rare catastrophes as the Krakatoa explosion. 



The closing paper was by Mr. Chandler, ' On 

 the use of the zenith-telescope for latitude.' 



PROCEEDINGS OF THE SECTION OF 

 BIOLOGY. 



The regular work of the biological section began 

 on Thursday, and a partial classification of the 

 papers into botanical and zoological added consid- 

 erably to the interest and convenience of those pres- 

 ent. Some have proposed a divison of the section 

 of biology into botanical and zoological sections, 

 but this, with a small meeting, seems hardly de- 

 sirable, as there are apt to be only enough papers 

 to occupy the time. 



Among the first of the botanical papers was one 

 by Prof. W. J. Beal. giving a comparison between 

 the hygroscopic cells of grasses and sedges. In 

 both grasses and sedges, as has long been known, 

 there are one or more longitudinal rows of cells 

 on each leaf, the function of which is to fold or 

 close the blade in times of drought, and thus pre- 

 vent too rapid evaporation of moisture froin the 

 surface. These rows of cells, as well as the cells 

 themselves, vary in shape, size, ,and distribution 

 in the different genera and species, and may have 

 some value . in the discrimination of critical 

 species. The most interesting point brought out 

 was, that many parallels exist between the genera 

 of grasses and sedges in the arrangement of these 

 hygroscopic, or, — as Professor Beal chooses to 

 term them, — bulliform cells. 



The paper of Messrs. J. M. Coulter and J. N. 

 Rose, giving a synopsis of the North American 

 Ijines, based on leaf-stnicture, had some pomts in 

 common with the one just mentioned, and was of 

 especial value from a systematic stand-point, from 

 the fact that any species in this somewhat difficult 

 group can at once be distinguished by the pecul- 

 iarities of its minute leaf -structure ; and the re- 

 sults of the author's observations are shown to be 

 worthy of attention from the fact that a classifi- 

 cation based on these characters is, in its broader 

 featiu-es, closely like that of the late Dr. Engel- 

 mann, which, as is well known, took into consid- 

 eration the whole tree. 



The relations of germs to disease naturally oc- 

 cupied a prominent place in the proceedings of the 

 section, and the presence of over half a dozen in- 

 vestigators in this line made the discussions inter- 

 esting. Dr. D. E. Salmon read two papers bearing 

 on the causes of immunity from a second at- 

 tact of germ diseases. There are three possible 

 explanations : 1*^, something is deposited in the 

 body during the attack which is unfavorable to 

 the germ ; 2°, something has been withdrawn 

 which is necessary to its development ; 3°, the 

 tissues have acquired such a tolerance for the 

 germ or for an accompanying poison that they are 

 no longer afi'ected by it. Dr. Salmon favored the 

 last view, and gave details of a large number of 

 experiments to substantiate his opinion. He said 

 that Metchinkoff's phagocyte theory was not 

 wiiolly satisfactory, and that large doses of the 

 germs were more powerful than small ones. He 

 attributed their action to a poison which was a 

 result of their growth, and thought that a large 

 dose had a greater effect because the poisons 

 benumbed or killed the cells, thus giving the 

 bacteria a better chance to grow and to thus pro- 

 duce more poison. 

 ' Dr. Joseph Jastrow gave an account of some 



