Kana can be modified with small marks that change the pronunciation of the consonant. The first of these is dakuten, also called ten-ten, which consists of two small strokes (゛) placed in the upper right of a character. Adding dakuten voices an otherwise unvoiced consonant. For example, か (ka) becomes が (ga), さ (sa) becomes ざ (za), た (ta) becomes だ (da), and は (ha) becomes ば (ba). The same transformation applies to every character in the K-, S-, T-, and H-rows, producing a full set of voiced consonants. In katakana, the voiced forms are written the same way: ガ for /ga/, ザ for /za/, ダ for /da/, バ for /ba/.
The second modifier is handakuten, also called maru, which is a small circle (゜) used exclusively on the H-row. Adding handakuten converts the "h" sound to a "p" sound, so は (ha) becomes ぱ (pa), ひ (hi) becomes ぴ (pi), ふ (fu) becomes ぷ (pu), へ (he) becomes ぺ (pe), and ほ (ho) becomes ぽ (po). The katakana half-voiced forms follow the same rule, giving パ for /pa/ and so on. There are no handakuten forms for any row other than the H-row, because Japanese does not use a "p" series for any other consonant.
One subtle point concerns the voiced form of ち (chi). Theoretically, adding dakuten to ち produces ヂ, pronounced /ji/. In practice, however, the standard character used for /ji/ is ジ, derived from the voicing of シ (shi). Learners will encounter ヂ only in rare compound words, while ジ appears constantly in everyday katakana loanwords. The same situation arises for the voiced form of つ, written ヅ and pronounced /zu/, although this character appears more often than ヂ. Recognizing these patterns helps learners read unfamiliar words more confidently.