Meiotic Nuclear Divisions of Galtonia candicans. 749 
considered to be parallel threads and parallel aggregations of linin in the 
somatic cells. Further, that these parallelisms are the condensations of the 
sides of somatic chromosomes, whether of portions only, or of whole chromo¬ 
somes. Parallelisms are also found in the heterotype prophases by most 
investigators, but these parallelisms are interpreted in three different ways. 
(1) Although the parallelisms in the heterotype prophases may have 
identically the same appearance as the parallelisms of the somatic divisions, 
yet that in the heterotype division each side of the parallelism has a 
different origin to the other side, and each represents a whole somatic 
chromosome. 
(2) That the parallelism in the heterotype prophases must be con¬ 
sidered homologous with the parallelism of the somatic prophases, and 
consequently that in the heterotype prophases, as well as in the somatic 
prophases, each side of the parallelism has the same origin, that is to say 
that it arises from the same longitudinally split chromosome. 
(3) That the parallelism found in the heterotype prophases has no real 
significance but is a mere coincidence. 
The views of those who hold that the parallelism in the heterotype 
prophases is the pairing of homologous chromosomes will be taken first. 
(1) Rosenberg ( 25 ) (1907) figured chromatin masses in the unreduced 
number in the heterotype prophases of Tanacetum vulgare which resembled 
those aggregations in the somatic divisions. These masses united in pairs 
in synapsis and each member of the ‘ Gamosomen-Paar ’ became a univalent 
chromosome. In Crepis virens , where the unreduced number of chromo¬ 
somes is six and the reduced number is three, there are six prochromosomes 
in the somatic divisions ; these six reappear in the early heterotype prophase 
and then unite in pairs, which, according to Rosenberg, must be regarded as 
the conjugation of two whole somatic chromosomes. 
Overton ( 24 ) (1909) states: £ I am very strongly convinced that the 
arrangement of the prochromosomes in somatic and young germ cells is the 
same, that is, they are parallel in pairs,’ and that c the homologous parental 
elements are therefore associated in pairs when they enter the reconstructive 
stages of the germ nuclei ’. 
The Schreiners ( 28 ) have published a beautiful figure (PI. 23, Fig. 10) 
of the ‘ Conjugation ’ of the chromosomes in the early heterotype prophase 
of Salaniandrci maciilosa which exactly simulates the condensation of the 
chromosome in the somatic nuclei of the roots of Galtonia (Pis. LIX and 
LX, Figs. 19 and 20). There is the same irregular and partly beaded 
approximating sides, joined to the skeleton chromosomes on either side by 
delicate connexions. 
Lundegardh’s ( 16 ) drawings of the early heterotype prophases of 
Calendtda officinalis closely resemble those of Galtonia , but he considers 
that the parallelisms represent the pairing of homologous chromosomes. 
