144 



KNOWLEDGE A: SCIENTIFIC NEWS. 



[JlLV, 1905. 



powers, 29 averaging 0.23 sun-powers, 19 averaging 

 .034 sun-powers, and no less than 28 averaging only 

 .005 sun-powers. The majority arc dwarfs : it i'- 

 quite as necessary to provide for dwarfs as for giants 

 in the Milky Way, but there appear to be very few 

 stars of less than the 17th magnitude ! The highest 

 authority on this point, the late Dr. Isaac Roberts, 

 writing in "Knowledge" (1901, page 11) about a 

 photograph of the Milky Way, which shows stars down 

 to the 17th magnitude, says, " The evidence of photo- 

 graphs stroni;ly indicates that those vacant places 

 which appear after exposures of 7 to 12 hours are 

 really void of stars, because exposures of only 90 

 minutes show the same stars down to the faintest 

 magnitudes." His remarks should be read in cxlensa, 

 as they are more conclusive than thus condensed. 



To enable my readers to weigh for themselves the 

 value of the evidence given here, in my first article and 

 in my " Ivoad-Book," 1 worked out a table which gi\es 

 the sun-power of stars of different magnitudes at differ- 

 ent distances. I think it comprehends all the .Milky 

 Way stars, but it can be readily extended up or down, 

 or right or left, e.g., 3rd m;ignitude, at 1,000 light- 

 years =6,800 sun powers; or 6th magnitude at ion 

 light-years = 4.37 sun-powers. 



From the evidence produced, my readers can judge 

 as well as I — some better — what are the probable 

 dimensions of our Stellar Universe. The first magni- 

 tude stars are exceptionally large, and if they only are 

 considered, and if Mr. Walkey's estimates be correct, 

 we might allow a radius of about 5,000 light-years for 

 the Milky Way, but if 80 per cent, of the stars are 

 smaller than the .Sun, the radius would not be more 

 than about 1,500 light-years. I have assumed in my 

 drawing the truth lies between these extremes. I 

 claim to be no aiithirity in such matters, but seek onl\ 

 to illustrate by my diagrams and stereograms such 

 data as I can collect, so that anyone c;m see what they 

 mean. The scale I discovered, which, taking the -Sun's 

 mean distance as one inch, makes the distance travelled 

 by light in one year one mile, is, I find, to most people, 

 a real help. Objects in space of three dimensions can 

 be drawn without distortion as well upon a flat surface 

 as upon a sphere. The excellent maps now being pub- 

 lished in " Knowledc.e " appear to have no distortion 

 if they are viewed with a lens so that the eye is 5^ 

 inches above the centre. It is greatly to be desired 

 that some owner of a clock-driven telescope would 

 take and publish a set of lantern slide star maps (which 

 could be enlarged as required), using a stigmatic lens of 



Dr. F'laston in an interesting article (" Knowledge," 

 1903, page 154) gives a sample of the Milky Way 

 (area Man part of sky). He has estimated magnitudes 

 (from 5.2 to 14) of the 1,761 stars shown therein. 



To help our judgment I worked out the sun-powers 

 ff)r these stars, upon the supposition they are either 

 1,500, or 3,000, or 7,500 light-years distant. They 

 come out as follows : — 



3I inches focus and short exposure, so ris not to go 

 much below the 6th magnitude. 



In my first article I suggested a model which would 

 show the Sun and stars of estimated parallax shining 

 with proportionate luminosities. I have constructed 

 such a model more efhcicntly than I had hoped, and 

 upon convenient tlal surfaces, in the six large stereo- 

 grams I recently exhibited in London. 1 may even say 

 that when 1 took the members of three learned societies 

 many hundred billion miles out into space and showed 

 them the .Sun and stars shining in their prr.pcr colours, 

 with their relative luminosities and hanging in space at 

 their eslimalcd dislaiircs, I broke through ihe crvstal- 

 line vault which iias so long imprisoned even those who 

 know it is but an illusir)n. 



Notice to Readers. 



We beg to remind regular subscribers that the Special Number, 

 to be published on July 15th, will run concurrently with the 

 usual numbers, the paging being continuous, and the matter 

 will be included in the index for the year. 



