1883.] 



MICROSCOPICAL JOURNAL. 



225 



are all but one authority, and that one 

 is Gulliver. 



Therefore, I claim that niicroscop- 

 ists should abandon the T5-._>\y^-inch as 

 a standard, and should determine a 

 new standard by means of the best 

 modern appliances, and I venture the 

 opinion that such a standard will be 

 found to b2 nearer to vfj^gjy-inch than 

 is generally suspected. But most 

 especially do I object to careful meas- 

 urements made at this late day being 

 compared with the ^oVii^ or any of 

 the old reported measurements, to de- 

 termine the character of the blood 

 measured. If a comparison is to be 

 made let each microscopist make his 

 own comparison directly, by measur- 

 ing the things to be compared with the 

 same apparatus, and this comparison 

 he can rely upon with full confidence, 

 while he can not so rely upon a com- 

 parison of his own measurement of 

 one with somebody else's measure- 

 ment of the other. 



Late measurements of several thou- 

 sand corpucles of various bloods 

 under various conditions in which I 

 have participated with a fellow-worker, 

 have convinced me of the danger of 

 reliance upon " authorities " in this 

 matter if a correct result is to be hoped 

 for. In a future number I may com- 

 municate some of the very interesting 

 facts observed. In the meantime, 

 as bearing upon the subject, and set- 

 ting forth fully my views of micrometry 

 in general in relation to minute objects, 

 including blood, I state the following 

 propositions as being generally and 

 with rare exceptions true, to wit : — 



1. Different observers, using the 

 same apparatus and viewing the same 

 object, will almost invariably differ 

 slightly in their measurements of it. 

 Between some observers the difference 

 will amount to nearly five per cent, of 

 the lesser measure, when using powers 

 of 1500 or upwards. 



2. The lower the magnifying power 

 used the greater will be this individual 

 error, in proportion to the actual size 

 of the object. 



3. Measurements made by one ob- 

 server, cannot, as a rule, be accurately 



compared with those of the same ob- 

 ject made independently by another 

 observer, even if similar apparatus and 

 powers are used by both ; while if 

 different apparatus and powers be 

 used the comparison becomes much 

 less satisfactory, and is practically of 

 but little value. 



4. Even when the errors of the re- 

 spective micrometers and apparatus 

 and the personal equation of the re- 

 spective observers are known, such 

 measurements are only susceptible of 

 comparison with approximate accu- 

 racy : and where these elements are 

 unknown, no comparison of sufficient 

 value to affect vital or important inter- 

 ests, is possible. 



5. Hence, measurements made with 

 the best modern objectives and ap- 

 paratus, and with powers of 1000 and 

 upwards are not susceptible of accurate 

 comparison with the older measure- 

 ments of the same objects, even where 

 powers approaching or exceeding 1000 

 were used for the old measurements. 



6. Therefore, with special reference 

 to blood-corpuscles, the conclusion is, 

 that until an identical standard of 

 measurement is used by ail observ- 

 ers, measurements of blood cor- 

 puscles made for the comparison of 

 different kinds of blood, should not 

 be stated or compared in terms of any 

 other measure, nor in comparison with 

 other (and especially the older and 

 more commonly published) measure- 

 ments of blood, but should be stated 

 and compared solely in terms of the 

 micrometer used, giving also the vari- 

 ation, if any, of this micrometer from 

 some generally accessible and reli- 

 able standard. By this means the 

 sources of error will be confined to a 

 single element, which will be constant 

 and proportionate for all objects 

 measured by the same observer under 

 the same conditions, and will, therefore, 

 afford a true comparison of the rela- 

 tive size of the objects so measured. 

 And by means of the tabular variation 

 given, any other observer can compare 

 the measurements so made with his own 

 or others' measurements equally as 

 well as the observer who reports them. 



