July 16, 1920] 



SCIENCE 



63 



cosmic and terrestrial physics. Continuous 

 and higKly refined observations on the sun 

 have demonstrated the variable nature of the 

 "solar constant." The so-called constants of 

 heat conductivity, diffusion, viscosity, etc., 

 pei-taining to the ocean also vary with the 

 conditions, though they are all thousands of 

 times larger than the corresponding labora- 

 tory values. 



For example, to determine the upwelling 

 velocity in the southern California coastal 

 region the author applied the classical equation^ 

 for the diffusion of salts in a medium moving 

 with the velocity W, to seasonal observations 

 of ocean salinities at a series of depths, and 

 obtained the value 40 in C.G.S. units for the 

 diffusion constant /t^, while a Norwegian in- 

 vestigator, Jacobsen, obtained values varying 

 from 0.3 to 11.4 for different regions of the 

 sea near Denmark. The laboratory value for 

 the diffusion coefficient of ocean salts in 

 water is only .0000125. The upwelling veloc- 

 ity in the southern California region was also 

 determined by applying to serial ocean tem- 

 peratures Fourier's equation for the flow of 

 heat in a moving medium. The conductivity 

 constant for this ocean region was found to 

 be 30, while the laboratory value of the coeffi- 

 cient of conductivity of sea-water is only 

 .0012. 



The values of such constants found under 

 the simple conditions of laboratory control 

 are known to depend upon the temperature of 

 the fluid. This is in turn an index of the 

 complex molecular activity. In the ocean, 

 the corresponding variable factor is the rate 

 of interchange of small parts of the water in 

 the ever present alternating couvective circu- 

 lation. 



Complicated as those phenomena are, en- 

 couraging results have already come from 

 quantitative studies, not only in oceanog- 



i Ocean, temperatures, their relation to solar 

 radiation: quantitajtive compaTlsona of certain em- 

 pdrioal results with those deduced by principles and 

 methods of mathemaiticaJ physics by George f. 

 MoEwen, 1919, Semicentennial Publications of the 

 University of OaHfornia, 1868-1918, pp. 336-421, 

 19 figs, in text. 



raphy, but also in other geophysical investi- 

 gations. George F. McEwen 

 ScBiPPS Institution foe 

 Biological Research, 

 University or California 



THE DIFFERENTIAL STAINING OF PLANT 

 PATHOGEN AND HOST 



The well-known difficulty experienced in 

 staining to differentiate pathogen from host 

 tissues in phytopathological studies needs no 

 comment. In order to obviate this difficiilty 

 the writer has tried numerous combinations 

 of stains and finally a method was hit upon 

 which gives uniformly satisfactory results 

 from the histological point of view. It is not 

 intended for cytological studies although even 

 for these there may be possibilities in the 

 method. 



The comparatively short time required to 

 complete preparations, and the fact that stu- 

 dents not yet expert in microtechnique can in 

 most cases obtain good mounts, decided the 

 question of publication. 



STAINS 



1. Magdala red. A 2 per cent, solution in 85 per 



cent alcohol. 



2. Licht grun. A 2 per cent, solution in clove oil 



to which has been added a few drops of abso- 

 lute alcohol. 



METHOD 



1. Dissolve paraffin in xylol and wash in absolute 



alcohol. 



2. Wash in 95 per cent, and 85 per cent, alcohols. 



3. Stain with Magdala red 5 to 10 minutes. 



4. Eemove surplus stain and wash in 95 per cent. 



alcohol. 



5. Stain with Licht griin in clove oil for 1 to 3 



minutes. 



6. "Wash in absolute alcohol, or in carbol-turpen- 



tine. 



7. Clear in xylol and mount in Canada balsam. 



The time factors may require slight modi- 

 fications in some cases but a microscopic ex- 

 amination of the slide will enable the worker 

 easily to determine the variation required. 

 As a rule the staining with Licht griin is very 

 rapid and if overstaining occurs the red be- 

 comes tinged with purple although this may 



