544 
Fraser and Brooks . — Further Studies on the 
prophase. According to a considerable body of workers the chromosomes 
become joined end to end, and the early duplication of the spireme is due 
to fission ; this view was elaborated by Farmer and Moore (5) in 1905 , and 
has again lately received confirmation in the work of R. R. Gates (10). 
Gates finds the premeiotic chromosomes of Oenothera ruhrinervis arranged 
in linear series and separated by delicate threads of linin. From this 
spireme the chromosomes break off in pairs, each pair forming one of the 
gemini which are thus half as numerous as the chromatin masses on the 
spireme. 
Other investigators, including Strasburger, Gregoire, and their pupils 
have regarded the paternal and maternal spiremes as arranged side by side 
in parallel lines ; the chromosomes are held to conjugate laterally, and 
to undergo a more or less intimate union during the so-called synaptic 
or first contraction. Notably Overton (15), studying the pollen mother- 
cells of Thalictrum purpurascens and some other species, has described 
a series of double bodies lying along the spireme; he regards these as 
gemini each made up of two premeiotic chromosomes lying side by side. 
It is irresistible to compare with such bodies the chromosomes figured 
by Gates, and at first sight they appear to be strictly comparable, but it is 
important that while Gates’s structures are said to be twice as numerous as 
the chromosomes of the homotype telophase, Overton’s are described as 
equal to these in number. In other words, the heterotype prophase of Oeno- 
thera shows 3 n chromosomes arranged in linear series and separating two 
and two, while in Thalictrum there appear to be two spiremes with n chromo- 
somes each, and these lie side by side. If both these accounts be correct it 
must be inferred that the chromosomes become associated in the one case end 
to end, and in the other laterally. Gates has further shown that in Oeno- 
thera hybrids the union of fourteen and seven chromosomes in fertilization 
gives rise to ten or eleven ‘ gemini ’ heterotype prophase. 
He points out that this could scarcely occur if the paternal and maternal 
spiremes lay side by side. On the other hand in Drosera (Rosenberg (16)) 
a form with twenty chromosomes crossed by one with ten gives thirty in the 
sporophyte, and twenty, ten large and ten small ( 10 + 10 '^, in meiosis. 
This is not inconsistent with lateral union, but there seems no reason 
why it should not equally be the result of association end to end, and it is 
suggestive in this connexion that the bivalent chromosomes are of greater 
length than the others. 
In the Discomycetes hitherto investigated the chromosomes were 
found to become paired end to end, and our present studies confirm this 
conclusion. In Mildews, on the other hand, Harper (12) has shown that the 
chromosomes are recognizable throughout the resting-stages, and remain 
