The aim of this research is to investigate, whether irregular light variations of late type giant pulsating variables can be understood as chaotic behaviour or not, oh the basis of observed light curves of long period variables including Mira, semiregular and irregular type stars, whose periods are about several hundred days.
In order to get basic data for analysis, I carried out photometric observations for long period variables and intended to make database, which collected unpublished visual observations for these stars in Japan in the past. To compare with long period variables, I also observed ZZ Psc, one of the pulsating white dwarf stars, whose period is very sort and about 1000 second.
Photometric observations are made at Dodaira station of National observatory and the database, which has about one million observations for about five thousand stars including long period variables from early 1910's to. the present, is completed. Since this database will be very useful to various study on variable stars, I intend to open wide to the researchers.
Using the data acquired as above, I made analyses statistically and dynamically to find chaos for some Mira and semiregular stars, which have enough data, by the methods of period analysis, making first-return maps, correlation integral and so on.
The results are as follows : In spite of many trials to search chaos in Mira and semiregular stars, it seems that the strict evidence of chaos cannot find. However, if the light curves are chaotic, the dimension of semiregular stars is larger than Miras. From the first-return maps, carbon stars show peculiar features compared with other spectrum types. For ZZ Psc, I cannot find chaos, too, but find period variations.