94 

 duct into another one, as Tison^ shows in Hamamelis virginiana . 



The size of the tyloses is greatly dependent ^on the 

 space at their disposal. They sometimes grow out into cut 

 sducts, and then attain a considerable size. I have seen them 

 grow to gigantic sacs on the cut surfaces of Platanus cuttings 

 and incompletely cover the exposed surface of the wood. 



flOl) Content and wall of the parenchyma cells, developing 

 tyloses, remain essentially unchanged in the case of tylose 

 formation. The n ucleus of the parenchyma cell does not divide 

 (compare the footnote) . Often it wanders over' into the tylose 

 usually the young tyloses seem to lack nuclei, since the 

 nucleus leaves the mother cell only in late stages of tylose 

 development. «i The membrane s of tyloses, as is true in many 

 other hypertrophies, are often very thinl In other cases, 

 massive thickenings of the walls may be observed and even cor- 

 responding bordered p;^ts on the contact surfaces of adjacent 

 i? °f ®? ^^^S* ^^'' ' Moller^ found very thick wall tyloses with 

 the habit of growth of siione soils (stone tyloses) in the 

 wood of P iratin era guianensis, Molisch (loc, cit.j). 273), in 

 Mespilod aphne Sassafras (compare fig. 32.) In the case of 

 Piratinera all the tyloses have become stone-cells, the ducts 

 being completely stopped up with them. By this means "Jfhe 

 homogeneity of the wood is significantly increased" (Molisch) . 

 In Mespilodaphne the stone- tyloses alternate with relatively 

 thin-walled ones, (compare the figure). The wood^-parenchyma 

 . cell3 of Mespilodaphne, developing tyloses, are otherwise 

 rather thin-v/alled. 



1. Rech. s. la chute d. feuilles chez 1, Dicotyl. These 

 Caen, 1900. 



2. Molisch has shown already (Zur Kenntnis d. Thyllen, nebst 

 Beob; u. W-undheilung in d. Pfl. Sitzungsber, Akad. Wiss. Wien, 

 1888, Bd. XCVII, Abt. I, p. 264), that the relations between 

 the groY^th of the cell wall and the position of the nucleus 



in the cel^, which Haberlandt proved for many cases (Ueb. die 

 Beziehungejl 3W. Funktion und Lage des Zellkerns, Jena, 1887) 

 are not always recognizable in, the fozmation of tyloses. Sim^ 

 ilar cases, however, are not lacking ijn which the nucleus at 

 first remains at a distance from the distensions of the cell, 

 produaed by jsuperficial growth of the walls and wanders over 

 into the appendage only when it is finished, as, for example, 

 in the haustoria of the Erysipheae, which obtain their nuclei 

 only after concluding growth. The neucleus forBBS itself in 

 to the lumen pf the haustorium through the thin neck of the 

 latter, (compare Smith, The Haustoria of the Erysipheae. Bot, 

 Gaz, 1900, Vol. XSIX, p. 153, 167); further in the young 

 basidia of the Basidiomycetes according to Maire (S. 1, cyto- 

 logie des Gastermmycetes. C. E, Acad. Sc. Pairs, 1900, T. EXXI, 

 p. 1346). According to K. Tamba (Die Herkuhft der Zellkerne in 

 den Gefassthyllen von Cucurbita Sitzungsber. Phys. Mediz, Soc. 

 Erlangen, 1887, Bd. XIX» p. 4) the nucleus divides at times in 

 the ' parenchyma cells of Cucurbita producing tyloses, one of the 

 daughter nuclei goes over to the tylose, the other remains in 

 the mother cell. Closer testing would be much desired. 



3. Rohstoffe des Tischler- und Drechslergewerbes. 1883, 

 Bd. I, p. 143. 



