HBr을 이용한 트렌치 식각시 식각 방지막의 형성과 이들이 결함 생성 및 분포에 미치는 영향을 고분해능 투과전자현미경을 이용하여 연구하였다. O2 및 다른 첨가 가스로 SiOxFy, SiOxBry 등의 식각 방지막을 표면에 형성시켜 벽면 undercut을 방지하고 표면의 거칠기를 감소할 수 있었으며, 이후의 트렌치 채움 공정에서 void 가 없는 잘 채원진 구조를 얻을 수 있었다. 형성된 식각 방지막은 격자 결함의 생성 및 이들의 분포에 영향을 미쳤다. 대부분의 식각 유도 결함들은 트렌치 바닥의 가장자리에서 10Å 이내의 깊이로 분포하였으며, 잔류막의 두께에 의존하였다. 두꺼운 잔류막층 아래로는 결함들이 거의 사라졌으며, 결함층의 깊이와 잔류막 두께는 대체로 반비례하는 것을 나타났다. 기판 내에 존재하는 결정학적인 결함들은 식각종의 입사각이나 에너지에 의존하는 반면에,식각된 표면에서 관찰되는 결함들은 트렌치 식각동안 형성되는 이러한 잔류막의 두께에 크게 의존하는 것으로 나타났다.

한국재래종과 도입종을 포함한 62계통의 고추를 경북대학교 부속농장에 재배하여 종자를 증식하는 동시에 개화소요일수를 포함한 16가지 형질의 특성을 조사 기록하였다. 개화소요일수에 있어서는 KC85(칠곡 동명)와 KC116(PI308791), KC154(Cherry Sweet) 및 KC185(조선고추)가 83일로 가장 조생이었고 KC139(Serrano Chili)와 KC103(PI297488)이 126일로 가장 만생이었다. 과실을 따는데 걸리는 힘에 있어서는 KC85(칠곡 동명)가 0.45kg으로 가장 쉽게 따지고 KC162(PI224444)가 3.34kg으로 가장 많은 힘이 걸렸다. 흰가루병에는 KC10(Pimiento), KC23(PI224423), KC51(PI257044), KC56(PI257053), KC126(PI358812), KC119(PI322719), KC143(Emerald Giant), KC148(VR-2), KC163(PI224445), KC166(Papari Mild)이 발병이 적었다. 탄저병 발생은 KC39(Serrano Chili)에서 가장 적었다. 고추 품종의 수집 및 도입활동을 계속하여 한국재래종 14점과 도입종 25점을 추가 확보하였다.

The study was conducted to indentify systematically current marketing structure and markeiing efficiency of major cash crops such as red peppers and garlics. That various data and information have been used to analyze the problems related, some of them have come from secondary sources, but major parts of real circum stances were primarily obtained from the scheduled field survey. Some of valuable facts and consequences stemmed from the research can be summarized as fellows: First of all, the commercial rates of peppers and garlics produced by sample farms were 87.5 percent and 69 percent, respectively. On the other hand, annual handling volumes per marketing firm of those two crops were on the average estimated into 51,000 kyun(about 600gr) for peppers and 20,000 hundred bulbs of garlic. As seen physical losses over the entire marketing channels of each crop, it was approximately calculated in 14.7 percent of peppers and 21.1 percent of garlic, respectively. Secondarily, it was found that marketing chanells of these crops are very diversified and complex. However, the five-staged chanell is typical: i.e., from producers to local consigners, to local carry-out merchants, to consigner-wholesales, and finally to retailers. Thirdly, it was also estimated that gross marketing margin, marketing cost and profit based on average grade of peppers were 56.1 percent, 22.6 and 33.5 percent, respectively, having 43.8 percent of farmer's share, and those of garlic 38.9, 17.1, 21.8 and 61.1 percent, separately. To the end, it seems that current collective bargaining power of farmers' cooperatives is virtually weak compared to free individual marketing firms in terms of bssiness volumes and quantity handled over the year.

Kim, Sang-Ki. 1998. The Middle Construction and Agentivity. Studies in Modern Grammar 13, 25-39. This paper has characterized the nature of agentivity in the middle construction. Evidence of various types suggests that although agents cannot be overtly expressed in English middles, they are present underlyingly. Instrumental with-phrases may appear in middle constructions. The presence of an instrument implies the existence of an agent to use it. Other arguments can be made for the presence of an agent in the middle. Phrases such as with (no, little, a little, some, a lot of) effort and (with, without) difficulty, for example, imply agentivity, since effort is energy expended by an agent. Middles have much in common with unaccusatives. Both types of predicate have one argument, a non-agent, which shows up as the subject in surface structure but is an underlying object. The difference between the middle and the unaccusative is that the former has an agent in underlying lexical semantic structure, while the latter has no agent at any stage in the derivation. I have used a variety of syntactic and semantic tests to differentiate middles from unaccusatives.