diff --git a/problems/kamacoder/0099.岛屿的数量深搜.md b/problems/kamacoder/0099.岛屿的数量深搜.md
index 06be9268..eed21eca 100644
--- a/problems/kamacoder/0099.岛屿的数量深搜.md
+++ b/problems/kamacoder/0099.岛屿的数量深搜.md
@@ -326,6 +326,75 @@ if __name__ == '__main__':
 
 ### Go
 
+我们使用一个visited数组，记录下那些位置被遍历过。分为两层遍历。初始化一个visited数组和原始的grid一样大，用来记录哪些陆地被遍历过
+
+第一层遍历遍历整个grid数组的元素，遇到陆地，就在对应的visited数组里标记，并且执行深度搜索，深搜的逻辑是把当前位置的4个方向上的位置，全部判断一遍看是不是陆地，如果是，则在这个位置上再执行深搜，达到递归深搜的效果。所以深搜函数里是递归的。
+
+```go
+package main
+
+import (
+	"fmt"
+)
+
+func visitIsland(grid [][]int) int {
+	row := len(grid)
+	if row == 0 {
+		return 0
+	}
+	visited := make([][]bool, row)
+	//go的这种初始化方式真的丑陋
+	for i := 0; i < row; i++ {
+		visited[i] = make([]bool, len(grid[0]))
+	}
+	ans := 0
+	for i := 0; i < row; i++ {
+		for j := 0; j < len(grid[0]); j++ {
+			if grid[i][j] == 1 && !visited[i][j] {
+				visited[i][j] = true
+				ans++
+				visitGrid(grid, visited, i, j)
+			}
+		}
+	}
+	return ans
+}
+
+func visitGrid(grid [][]int, visited [][]bool, x int, y int) {
+	diff := [4][2]int{{1, 0}, {-1, 0}, {0, 1}, {0, -1}}
+	for _, arr := range diff {
+		nextX := x + arr[0]
+		nextY := y + arr[1]
+		if nextX < 0 || nextX >= len(grid) || nextY < 0 || nextY >= len(grid[0]) {
+			continue
+		}
+		if !visited[nextX][nextY] && grid[nextX][nextY] == 1 {
+			visited[nextX][nextY] = true
+			visitGrid(grid, visited, nextX, nextY)
+		}
+	}
+}
+
+func main() {
+	var row, col int
+	fmt.Scan(&row, &col)
+	if row <=0 || col <=0 {
+		return
+	}
+	grid := make([][]int, row)
+	for i := 0; i < row; i++ {
+		grid[i] = make([]int, col)
+	}
+	for i := 0; i < row; i++ {
+		for j := 0; j < col; j++ {
+			fmt.Scan(&grid[i][j])
+		}
+	}
+	//这里必须要打印，不然报错会显示潜在的数组越界
+	fmt.Println(visitIsland(grid))
+}
+```
+
 ### Rust
 
 ### JavaScript