Kusano, Further studies on Aeginetia indica. 
289 
appear outside tlie testa at the micropylar end of the seed (Figs. 
3, 4). These are highly togescent with abundant cell-sap. 
The nuclei are large and conspicnous, and tbe cytoplasm radiale 
fi’om tbem. At an advanced stage tbe globular cells increase in 
number generally up to 15 approximately (Fig. 12). As can be 
seen in Figs. 1, 2 and 6, tbese are not a new tissue, but only 
tbe epidermal cells of tbe radicle, but swollen up to nearly 4 times 
tbe original diameter. Simultaneously witb tbis cbanges aU tbe 
otber cells sweU up more or less inaking tbe embryo mucb larger 
in size; and judging from tbe number of cells seen in an optical 
section of tbe embryo before and after germination (compare Figs. 
2 and 6), it is very improbable tbat a multiplication of cells may 
be partly concerned in tbe increase of tbe size. An accumilation 
of starcb more especiaUy in tbe tissue under tbe globular ceUs is 
perbaps connected not witb tbe cell-diyision in tbis place, but witb 
tbe furtber development of tbe globular ceUs. 
Now foUows tbe outgrowtb of tbe globular cells one by one. 
Tbeir external wall protrudes so as to make tbem fii-st conical and 
tben papiUalike in form (Figs. 7, 9, 10). Tbe outgrowtbs proceed 
fui’tber until tbey become slender bairs gi’owing at times up to 1 
mm in lengtb. Tbe diameter of tbe bairs is mucb smaUer tban 
tbat of tbe globular cells, measuring 38 jj. on tbe average wbile 
tbe latter measui’e generally 115 ^ in diameter. Altbougb tbey 
belpng morpbologically to tbe category of trieb omes, yet tbey are 
not identical in structure and even in function witb tbe typical 
rootbairs (Schwarz, 1883); tbey are often septate or even branebed 
(Figs. 8, 9, 14), resembbng ratber tbe rbizoids of some cryptogamic 
plants (HabeiTandt, 1904, p. 200). If undisturbed, tbey are all 
straigbt and radiate from tbe radicular end in all dfrections as 
sbown in Fig. 8, but if one of tbem during its fuilber Prolongation 
sbould come in contact by its tip witb a young bost-root, it seems 
to attacb ^ itself firmly to tbe latter and tben to coil or contract 
tbrougb its wbole lengtb, wbereby tbe seedling is di'awn closer 
to tbe bost (Fig. 10). Tbis is evidently an advantageous contrivance 
for tbe parasite to facilitate its organic connection witb tbe bost, 
tbat is to say, tbe formation of tbe primary baustorium. In Fig. 
9 is sbown one of tbe bairs just adbering to a bost-root, and about 
to bend itself, wbile in Figs. 10 and 12 are sbown bafrs in a mucb 
contracted condition witb tbe radicular end brougbt mucb nearer 
to tbe bost. 
By wbat means tbe tip of tbe bafr fixes itself to tbe bost 
bas not yet been made out exactly. Tt is not impossible tbat a 
cementing substance is secreted by tbe bair, but tbere bas actually 
come under my Observation such a case as sbown in Figs. 9 and 
10, wbere tbe fixation was effected by a sbgbt penetration of tbe 
tip of a bair between tbe epidermal cells. My observations, bow- 
ever, are not extended enougb to justify tbe conclusion tbat tbis 
is a general case witb Aeginetia. 
So far as I know, such an organ bas not bitberto been 
described in pbanerogamic parasites. Analogons but not bomo- 
Beihefte Bot. Centralbl. Bd. XXIV. Abt. I. Heft 3. 19 
