— 16 — 



limited to a certain locality. In bad years, hardly a single seed is obtain- 

 able. Neither can the gap be filled up by the use of cuttings nor leyers 

 as this conifer cannot be propagated by these ways. Thus, often the 

 intending planters of Karamatsu forests are obliged to wait for the coming 

 of a good crop of seeds and must put of the execution of their plan for 

 several years. This causes loss of money and often induces dishonest 

 seedsmen to dispose of their stale stock of the seed at a high price. It 

 is necessary therefore to investigate how to so preserve seeds that a good 

 supply of vigorous seeds can be had at any time. This was the object of 

 the researches described herein. 



I. First Experiment 



The experiments were conducted in a space of four consecutive years 

 from March, 1906, to September, 1909. Seeds used for the purpose were 

 harvested in 1905 being those of Sugi, Kuromatsu, Karamatsu and Kusu. 

 They were subjected to different treatment as follows: — 



A. Preservation in an ordinary Room. 



1. Air-dried seeds in cotton sacked were hung up in an ordinary 

 airy room. 



B. Preservation in a Cellar. 



2. Air-dried seeds. 



3. Fire-dried seeds. 



4. Fire-dried seeds kept in small bottles hermetically sealed from 

 the air. 



As may be supposed, the first mentioned method is a form of 

 preservation of forest seeds most common in this country. 



In the so-called "cellar conservation," B, seeds were kept in a 

 cellar that was digged on the side of a slope found in the compound of 

 the Station. It is 8 m deep sidewise and is located 5 m beneath the 

 surface of the ground. The height of the cavity is 2 m and is deprived 

 of light inside. It has an almost constant temperature averaging 15.5°C 

 the year round. According to the data observed for 1906 and 1907 the 

 difference of the maximum and minimum temperature does not exceed 

 1°C. a year. The moisture inside the cellar is always at the saturating 

 point as is shown by the frequent condensation on the glass bottle con- 



