182 PROCEEDINGS OF THE ACADEMY OF [Feb., 



these Korschelt distinguishes a "Prereduction," where the first mitosis 

 is the reductional one, from a "Postreduction," where the second one is 

 reductional. Until quite recently I held that it was not of great 

 physiological importance which of these mitoses was reductional, pro- 

 vided that in all cases one of them was. But my own studies, extended 

 over a diversity of objects, have convinced me that very probably there 

 maintains a uniformity here also, even though the understanding of it 

 may not be immediately forthcoming. For surely out of the endless 

 diversity in small details a larger uniformity is gradually showing itself, 

 and as scientific thinkers it is our object to discover the uniformity. 

 The minutiae of phenomena are but stepping-stones, and too often 

 slippery ones, toward simple and broad concepts. By analogy with 

 other natural phenomena we should a 'priori expect uniformity rather 

 than diversity. From this standpoint I enter upon the discussion 

 again, with the conviction that all maturation plans must be either 

 prereductional or all postreductional, and that there can be no com- 

 promise. In our decision we must argue from the facts presented in 

 the early growth period of the ovocytes and spermatocytes, from their 

 first mode of formation and axial relations to each other, and not from 

 analyses of their definitive forms. Other things being equal, such 

 evidence must have the most weight which considers in the greatest 

 detail the full sequence of stages, and most particularly such as treats 

 minutely the early growth period. Too often follows upon a short and 

 incomplete series of observations a long discussion in print of possibili- 

 ties and probabilities, like shuffling with an incomplete pack of cards, 

 instead of an attempt to settle the matter with the microscope. 



The foundation of the argument for postreduction is in the work of 

 Hacker and Riickert upon the Copepoda, and with but few exceptions 

 this has been allowed to go unchallenged. On that account the facts 

 of these workers call for careful examination. 



Riickert (1894) studied the ovogenesis in three genera from the stage 

 of the young ovocyte up to the monaster stage of the first maturation 

 mitosis. From the number of chromosomes present in young blasto- 

 meres of one of these, he concluded that each chromosome of the ovo- 

 cyte is bivalent, eciual to two. He described most fully Cyclops 

 strenuus, so that we will examine his work upon this species. His figs. 

 4-8 present stages of the earlier growth period, by no means a complete 

 series. The first ovocyte in these stages has eleven chromosomes, each 

 in the form of two more or less closely apposed rods. Riickert assumes 

 that each of these is a longitudinally split chromosome; but while he 

 correctly assumes that each is bivalent, he does not determine the 



